Search

Based on the improved YOLOv8n, a steel plate defect detection and recognition method is proposed to address the high labor costs and workload of traditional tasks. SPPFELAN processes inputs in parallel to enhance computational efficiency by executing multiple pooling operations simultaneously. The parallel feature fusion module PscSE, using a mixed‐dimension SE attention mechanism (scSE), captures global and channel‐related information better, improving characterization capability. The EIOU loss function addresses the ambiguous aspect ratio definition of CIOU loss, enhancing detection accuracy and accelerating convergence. Results show the YOLOv8n‐PscSE‐SPPFELAN model achieves 76.9% mAP@0.5 on the Northeastern University steel plate dataset, a 4.6% improvement over the original YOLOv8n, with a computation amount of 7.7 GFLOPs, reducing resource usage and greatly improving detection speed. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Beijing University of Civil Engineering & Architecture is the property of Journal of Beijing University of Civil Engineering & Architecture Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Wearing safety helmets at construction sites is a major measure to prevent safety accidents, so it is essential to supervise and ensure that workers wear safety helmets. This requires a high degree of real-time performance. We improved the network structure based on YOLOv7. To enhance real-time performance, we introduced GhostModule after comparing various modules to create a new efficient structure that generates more feature mappings with fewer linear operations. SE blocks were introduced after comparing several attention mechanisms to highlight important information in the image. The EIOU loss function was introduced to speed up the convergence of the model. Eventually, we constructed the efficient model EGS-YOLO. EGS-YOLO achieves a mAP of 91.1%, 0.2% higher than YOLOv7, and the inference time is 13.3% faster than YOLOv7 at 3.9 ms (RTX 3090). The parameters and computational complexity are reduced by 37.3% and 33.8%, respectively. The enhanced real-time performance while maintaining the original high precision can meet actual detection requirements. [ABSTRACT FROM AUTHOR]

Considering the defect detection issues in electroluminescence (EL) of photovoltaic (PV) cell systems, lots of factors result in performance degradation, including defect diversity, data imbalance, scale difference, etc. Focal-EIoU loss, an effective defect detection solution for EL, is proposed based on the improved YOLOv5. Firstly, by analyzing the detection background and scale characteristics of EL defects, a binary classification is carried out in the system. Subsequently, a cascade detection network based on YOLOv5 is designed to further extract features from the binary-classified defects. The defect localization and classification are achieved in this way. To address the problem of imbalanced defect samples, a loss function is designed based on EIoU and Focal-F1 Loss. Experimental results are illustrated to show the effectiveness. Compared with the existing CNN-based deep learning approaches, the proposed focal loss calculation-based method can effectively improve the performance of handling sample imbalance. Moreover, in the detection of 12 types of defects, the Yolov5 algorithms can always obtain higher MAP (mean average precision) even with different parameter levels (Yolov5m: 0.791 vs. 0.857, Yolov5l: 0.798 vs. 0.862, Yolov5x: 0.802 vs. 0.867, Yolov5s: 0.793 vs. 0.865). [ABSTRACT FROM AUTHOR]

Copyright of Journal of Henan University of Technology Natural Science Edition is the property of Henan University of Technology Journal Editorial Department and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The cocoa tree is prone to diverse diseases such as stem borer, stem canker, swollen shot, and root rot disease which impedes high yield. Early disease detection is a critical component of diverse management processes that are implemented throughout the life cycle of cocoa plants. Consequently, several studies on the application of detection techniques to recognize diseases have been proposed by several researchers. This study proposes the YOLOv5m network for cocoa tree disease detection. The development of cocoa disease detection systems will aid farmers in early identification prompt response, and efficient management of related cocoa tree diseases which will ultimately increase yield and sustainability. To improve the performance of the YOLOv5m network, a Swin Transformer (Swin-T) was added to the backbone network to improve cocoa tree disease detection accuracy. By obtaining global information, the K-means++ algorithm was added to modify the choice of initial clustering locations, and Efficient Intersection over Union Loss (EIoU) was used as a bounding box regression loss function to speed up the bounding box regression rate, resulting in higher precision of the YOLOv5m network. The experimental assessment outcome of this study showed that the proposed method YOLOv5m (Swin-T, K-means++, EIoU) achieved 96% precision, mAP of 92%, and recall of 94%. Compared to the original YOLOv5m, precision improved by 5%, mAP improved by 6%, and recall by 5%. Comparing the proposed method to the conventional YOLOv5m, the latter showed improved performance and better accuracy with a high detection speed and compactness. This improvement offers a useful and effective method for detecting diseases related to cocoa trees.

Summary: Camellia oleifera typically thrives in unstructured environments, making the identification of its trunks crucial for advancing agricultural robots towards modernization and sustainability. Traditional target detection algorithms, however, fall short in accurately identifying Camellia oleifera trunks, especially in scenarios characterized by small targets and poor lighting. This article introduces an enhanced trunk detection algorithm for Camellia oleifera based on an improved YOLOv7 model. This model incorporates dynamic snake convolution instead of standard convolutions to bolster its feature extraction capabilities. It integrates more contextual information, thus enhancing the model's generalization ability across various scenes. Additionally, coordinate attention is introduced to refine the model's spatial feature representation, amplifying the network's focus on essential target region features, which in turn boosts detection accuracy and robustness. This feature selectively strengthens response levels across different channels, prioritizing key attributes for classification and localization. Moreover, the original coordinate loss function of YOLOv7 is replaced with EIoU loss, further enhancing the model's robustness and convergence speed. Experimental results demonstrate a recall rate of 96%, a mean average precision (mAP) of 87.9%, an F1 score of 0.87, and a detection speed of 18 milliseconds per frame. When compared with other models like Faster‐RCNN, YOLOv3, ScaledYOLOv4, YOLOv5, and the original YOLOv7, our improved model shows mAP increases of 8.1%, 7.0%, 7.5%, and 6.6% respectively. Occupying only 70.8 MB, our model requires 9.8 MB less memory than the original YOLOv7. This model not only achieves high accuracy and detection efficiency but is also easily deployable on mobile devices, providing a robust foundation for future intelligent harvesting technologies. [ABSTRACT FROM AUTHOR]

Due to the limited computational resources of portable devices, target detection models for drone detection face challenges in real-time deployment. To enhance the detection efficiency of low, slow, and small unmanned aerial vehicles (UAVs), this study introduces an efficient drone detection model based on YOLOv5s (EDU-YOLO), incorporating lightweight feature extraction and balanced feature fusion modules. The model employs the ShuffleNetV2 network and coordinate attention mechanisms to construct a lightweight backbone network, significantly reducing the number of model parameters. It also utilizes a bidirectional feature pyramid network and ghost convolutions to build a balanced neck network, enriching the model's representational capacity. Additionally, a new loss function, EloU, replaces CIoU to improve the model's positioning accuracy and accelerate network convergence. Experimental results indicate that, compared to the YOLOv5s algorithm, our model only experiences a minimal decrease in mAP by 1.1%, while reducing GFLOPs from 16.0 to 2.2 and increasing FPS from 153 to 188. This provides a substantial foundation for networked optoelectronic detection of UAVs and similar slow-moving aerial targets, expanding the defensive perimeter and enabling earlier warnings. [ABSTRACT FROM AUTHOR]

Safety equipment detection is an important application of object detection, receiving widespread attention in fields such as smart construction sites and video surveillance. Significant progress has been made in object detection due to the rapid development of deep learning. Multi-scale targets and complex scenes increase the likelihood of false positives and missed detections, which can affect the accuracy of the detection. To address this issue, this study proposes YOLOv7-DSE. It is a small complex target scene detection network based on the improved YOLOv7. Also, we have created a private dataset of safety equipment. First, we enhanced the ELAN and MP backbone networks. Backbone is replaced by ordinary convolution by the depthwise separable convolution. We enabled the backbone network to extract deeper image features without increasing the amount of parameters and computation. Simultaneously, the model incorporates the EIOU loss function to improve its convergence speed and positioning effect. Secondly, we propose a new ELAN-SPD structure in the head network. Based on the ELAN structure, a space-to-depth convolutional layer is added to fully downsample the feature map, preserving all learnable features. Our network model can better detect objects with significant size differences faced with complex scenes. YOLOv7-DSE achieved the mAP of 82.38%, surpassing the original YOLOv7 with 2.64%. The YOLOv7-DSE model has a minor size compared to the baseline model. Our improvement has reduced the model parameters by 22.4%. [ABSTRACT FROM AUTHOR]

With the development of autonomous vehicles and intelligent transportation, more accurate detection of pedestrians. However, pedestrian detection suffers from occlusion and small target. First, the HorNet to improve the higher-order spatial interaction capability of the model, expand the effective sensory field, and enhance the feature extraction of pedestrians. Then, ODConv to gain the variability of each dimension and capture rich information. Finally, a layer to increase the accuracy of detecting pedestrians at small scales. We optimize the regression prediction of the anchor using the Efficient IOU Loss (EIOU) function. Experimental data show that the mean average precision (mAP) of the HOD-YOLOv5 model achieves 83.5%, compared to YOLOv5, which is 4.4% higher than original YOLOv5, and the recognition speed of the HOD-YOLOV5 reaches 106.7 frames per second (FPS). This demonstrates that the proposed model could realize real-time pedestrian detection at a relatively small cost, which satisfies the requirements of uncrewed and intelligent transportation.

Autonomous driving is a prominent research area, with the primary challenge being the precise perception and interpretation of the surrounding environment. To address issues such as low resolution, dense occlusion, and small targets in complex road scenes, we propose an improved road target detection algorithm, SNCE-YOLO, based on YOLOv5. Firstly, to avoid the loss of feature information during training for low-resolution and small targets, the Space-to-Depth Convolution(SPD-Conv) is introduced to replace the original convolutional layer for retaining global feature information. Secondly, to address the issue of confusing complex backgrounds with detection targets when features are extracted too early and deeply, the Normalization-based Attention Module (NAM) is incorporated into the backbone network without increasing the computational effort, focusing on the channel and spatial information of interest. Thirdly, to enhance the capture and fusion of multi-level feature information, we optimized the neck network’s feature fusion module by implementing the C2f structure. Finally, to reduce the bias in the regression of the occluded target positions, we introduced the EIOU loss function to better define the relationship between the predicted and real frames. The SNCE-YOLO algorithm achieves a mean Average Precision (mAP@0.5) of 91.9% on the KITTI dataset, which represents a 2.6% improvement over the baseline model. Additionally, with a detection speed of 95.24 frames per second (FPS), it meets the requirements for real-time target detection in complex road scenes. Comparisons are also made with other mainstream target detection algorithms to verify the effectiveness of SNCE-YOLO.

The rapid identification of landslides in mountainous terrain through remote sensing imagery after earthquakes or heavy rainfall is crucial for assessing the severity of the situation and providing timely assistance for post-disaster rescue operations. However, traditional manual interpretation methods and computer vision techniques tend to be inefficient and time-consuming. To address this problem, we propose a novel algorithm called LSI-YOLOv8, which builds upon the lightweight YOLOv8 to enable fast and accurate landslide identification using remote sensing imagery. In this study, we introduce optimization techniques integrated into the YOLOv8 model. Firstly, we incorporate an expandable residual attention Dilation-wise Residual Segmentation (DWRSeg) module to reconstruct its neck structure, reducing parameters and computational load. Additionally, we enhance the positioning ability of the model by replacing the bounding box loss function Complete Intersection over Union (CIoU) with Efficient Intersection over Union (EIoU), optimizing it to improve computing speed. Through comparative experiments with other state-of-the-art models, the accuracy of our LSI-YOLOv8 significantly outperforms Mask-RCNN, YOLOv5, YOLOv7, YOLOv8, and YOLOv9 model, with mAP@0.5 being 9.0%,7.7%, 4.0%, 6.3%, and 2.9% higher than other models, while mAP@0.5 :0.95 is 3.1%,3.6%, 2.7%, 4.8%, and 0.3% higher than others, respectively. Furthermore, the detecting speed of the LSI-YOLOv8 model is 73.2 f/s, which is notably faster by 15.5 f/s, 3.5 f/s, 5.1 f/s, 5.9 f/s, and 4.3 f/s respectively, compared to Mask-RCNN, YOLOv5, YOLOv7, YOLOv8, and YOLOv9. The proposed method provides valuable insights for advancing landslide identification methodologies. The focus of subsequent work will be more on multi-source data fusion, such as Digital Elevation Model (DEM) data and Synthetic Aperture Radar (SAR) data to remove the influence of some background information to improve the accuracy of landslide identification.

Considering steel as one of the most widely utilized materials, the detection of defects on its surface has always been a paramount area of research. Traditional target detection algorithms often face challenges such as low detection accuracy, missed and false detections, insufficient feature extraction capabilities, and inadequate feature fusion in tasks related to steel surface defect detection. To address these issues, this study proposes an enhanced algorithm, YOLOv8n-SDEC, utilizing the open-source dataset NEU-DET from Northeastern University as the sample dataset. Initially, the study improves the original SPPF module to the SPPCSPC module, enabling the network to better emphasize the features of the target. Furthermore, to augment the network’s feature extraction capability, a fusion with deformable convolution is introduced, enhancing the extraction of features from defective targets. The traditional CIoU loss function is substituted with the EIoU loss function in YOLOv8n aiming to minimize the discrepancies in height and width between predicted boxes and ground truth boxes. This substitution is intended to hasten model convergence and improve localization performance. Lastly, CARAFE is employed to replace the nearest neighbor algorithm, reducing the loss of feature information due to upsampling operations. Experimental outcomes reveal that the accuracy of the enhanced model reaches 76.7%, marking a 3.3% increase over the traditional model. Compared to conventional steel surface defect detection algorithms, the algorithm introduced in this study achieves more precise detection of steel surface defects.

The complex environmental influences often complicate the detection of longitudinal tears in conveyor belts, resulting in insufficient detection accuracy, overlooked detection, and elevated false detection rates. In this study, we propose a new depth learning method specifically designed for detecting longitudinal tears in conveyor belts. This method employs a linear Charge-Coupled Device (CCD) camera to capture images of the conveyor belt. These images are subsequently processed with a modified version of You Only Look Once (YOLO)v7 model to identify instances of longitudinal tearing. The modified YOLOv7 model features Efficient Intersection over Union (EIoU) loss function as a substitute for the original loss function. Furthermore, a Simple Parameter-Free Attention Module (SimAM) is introduced in the detection head to improve detection accuracy. In this method, we introduced the SimSPPFCSPC module as a new spatial pyramid pooling model. This module enhances detection speed while maintaining detection accuracy. Experiment results demonstrate the effectiveness of the proposed method, achieving an impressive precision of 94.6% and a detection speed of approximately 110 Frames Per Second (FPS). Such accuracy and speed meet the requirements for online detection of longitudinal tearing in belt conveyors.

As technology continues to advance, the automation of strawberry production and picking is an inevitable trend. To address the pressing issues of insufficient detection of smaller strawberries and misdetection resulting from the intricate background of strawberry images, an improved strawberry recognition method based on the YOLOv8n model was proposed. The improvements are as follows: 1) The deletion of the $20\times20$ pixel detection layer with a focus on small target strawberries and the addition of a $160\times160$ pixel small target detection layer were implemented to reduce the model volume and enhance the network reconstruction. 2) In the neck portion, a global attention mechanism was incorporated. 3) The SPD-Conv method was applied to improve the detection capability of small taget strawberries. 4) To address the limitations of the CIOU loss function, the EIOU loss function was utilized. The results of the experiment conducted on the self-made strawberry dataset demonstrated that the volume of the improved algorithm model was reduced by 59.7%, its precision was improved by 1.3%, and its recall rate increased by 2.1%. Additionally, the mAP was enhanced by 1.6%. The detection time for a single strawberry fruit image on a GPU was 17. 2 ms, which rendered the improved model suitable for practical applications. The model was verified in terms of small targets, and it achieved better detection performance than yolov5n, yolov6n, and yolov8s. The proposed algorithm demonstrated improved detection capabilities, reduced model size, and better target detection of strawberries.

UAV aerial photos tend to have complicated backgrounds and dense targets that vary in size. Applying existing object detection algorithms to such images is often inaccurate and prone to misdetection and omission. To better improve the detection performance of UAV aerial photography, we proposed an improved small-target detection algorithm based on YOLOv5s: 1) We reconstructed the feature fusion network by introducing an upsampling layer, increasing the model’s focus on features from small targets and improving related detection accuracy. 2) We introduced the SPD convolutional building block to downsample the feature map without losing learning information, improving the model’s feature extraction ability. 3) We replaced the CIoU Loss function of the original model with EIoU to reduce the location loss during training and improve the regression accuracy. We experimented with the improved algorithm on the VisDrone2019 dataset and achieved mAP@0.5 of 44%, demonstrating a 10.7% improvement from the original model. The detection speed also increases to 99 FPS, indicating that the improved algorithm can maintain its real-time performance while improving its accuracy.

The YOLOv4 network is widely used in object detection tasks as a representative network, but there is also the problem that the complexity of the network model affects the detection speed. In this paper, we propose an improved MV2_S_YE object detection algorithm based on the YOLOv4 network to improve the detection accuracy while increasing the road object detection speed. Firstly, the backbone network CSPDarknet53 of the YOLOv4 network is replaced by the Mobilenetv2 network to reduce the number of parameters of the network; secondly, the channel attention mechanism is introduced, and the SENet module is embedded in the structure of the PANet to optimize the object detection accuracy; finally, the EIOU loss function is used to replace the CIOU loss function to improve the object detection accuracy further. The MV2_S_YE network is obtained and tested on Pascal VOC, Udacity, and KAIST datasets. To evaluate our approach, we compared MV2-S-YE with YOLOv4, YOLOv4-tiny, YOLOv7-tiny and YOLOv8s. The results show that MV2-S-YE 的 mAP@0.5 achieves 80.9%, 66.7%, and 94.8% on the VOC2007, Udacity, and KAIST test sets, respectively, and is higher than YOLOv8s on both the Udacity and KAIST test sets. On the VOC2007 test set MV2-S-YE achieves a detection speed of 45FPS which is higher than YOLOv8s.

Wearing safety helmets at construction sites is a major measure to prevent safety accidents, so it is essential to supervise and ensure that workers wear safety helmets. This requires a high degree of real-time performance. We improved the network structure based on YOLOv7. To enhance real-time performance, we introduced GhostModule after comparing various modules to create a new efficient structure that generates more feature mappings with fewer linear operations. SE blocks were introduced after comparing several attention mechanisms to highlight important information in the image. The EIOU loss function was introduced to speed up the convergence of the model. Eventually, we constructed the efficient model EGS-YOLO. EGS-YOLO achieves a mAP of 91.1%, 0.2% higher than YOLOv7, and the inference time is 13.3% faster than YOLOv7 at 3.9 ms (RTX 3090). The parameters and computational complexity are reduced by 37.3% and 33.8%, respectively. The enhanced real-time performance while maintaining the original high precision can meet actual detection requirements.

In order to achieve the safety of autonomous driving, it is crucial to accurately detect traffic signs. Addressing the issue of low accuracy in traditional methods for traffic sign detection, a modified YOLOv5 algorithm incorporating the RepVGG module for traffic sign recognition was proposed. Firstly, by replacing certain CBS modules in the original algorithm with the RepVGG module, the feature extraction capability was enhanced. Additionally, the channel block attention module(CBAM) attention mechanism was introduced in the feature fusion module to strengthen the detection model's resistance to interference. Finally, during network training, the efficient-IoU(EIoU) loss function was utilized to achieve more precise target positioning, thereby improving the algorithm's detection accuracy and iteration speed. Experimental results demonstrate that the improved YOLOv5 algorithm exhibits faster iteration speed. On the CCTSDB traffic sign dataset, it achieves improvements of 4. 99% in accuracy, 3. 62% in recall rate, and 1. 73% in average accuracy compared to the original YOLOv5 algorithm, making it more suitable for practical applications. [ABSTRACT FROM AUTHOR]

Targeting the issue that the traditional target detection method has a high missing rate of minor target defects in the lithium battery electrode defect detection, this paper proposes an improved and optimized battery electrode defect detection model based on YOLOv8. Firstly, the lightweight GhostCony is used to replace the standard convolution, and the GhostC2f module is designed to replace part of the C2f, which reduces model computation and improves feature expression performance. Then, the coordinate attention (CA) module is incorporated into the neck network, amplifying the feature extraction efficiency of the improved model. Finally, the EIoU loss function is employed to swap out the initial YOLOv8 loss function, which improves the regression performance of the network. The empirical findings demonstrate that the enhanced model exhibits increments in crucial performance metrics relative to the original model: the precision rate is elevated by 2.4%, the recall rate by 2.3%, and the mean average precision (mAP) by 1.4%. The enhanced model demonstrates a marked enhancement in the frames per second (FPS) detection rate, significantly outperforming other comparative models. This evidence indicates that the enhanced model aligns well with the requirements of industrial development, demonstrating substantial practical value in industrial applications. [ABSTRACT FROM AUTHOR]

Aiming to quickly and accurately detect abnormal behaviors of workers in central control rooms, such as playing mobile phone and sleeping, an abnormal behavior detection algorithm based on improved YOLOv5 is proposed. The technique uses SRGAN to reconstruct the input image to improve the resolution and enhance the detailed information. Then, the MnasNet is introduced to replace the backbone feature extraction network of the original YOLOv5, which could achieve the lightweight of the model. Moreover, the detection accuracy of the whole network is enhanced by adding the ECA-Net attention mechanism into the feature fusion network structure of YOLOv5 and modifying the loss function as EIOU. The experimental results in the custom dataset show that compared with the original YOLOv5 algorithm, the algorithm proposed in this paper improves the detection speed to 75.50 frames/s under the condition of high detection accuracy, which meets the requirements of real-time detection. Meanwhile, compared with other mainstream behavior detection algorithms, this algorithm also shows better detection performance. [ABSTRACT FROM AUTHOR]

To address the challenge of achieving a balance between efficiency and performance in steel surface defect detection, this paper presents a novel algorithm that enhances the YOLOv5 defect detection model. The enhancement process begins by employing the K-means++ algorithm to fine-tune the location of the prior anchor boxes, improving the matching process. Subsequently, the loss function is transitioned from generalized intersection over union (GIOU) to efficient intersection over union (EIOU) to mitigate the former's degeneration issues. To minimize information loss, Carafe upsampling replaces traditional upsampling techniques. Lastly, the squeeze and excitation networks (SE-Net) module is incorporated to augment the model's sensitivity to channel features. Experimental evaluations conducted on a public defect dataset reveal that the proposed method elevates the mean average precision (mAP) by seven percentage points compared to the original YOLOv5 model, achieving an mAP of 83.3%. Furthermore, our model's size is significantly reduced compared to other advanced algorithms, while maintaining a processing speed of 47 frames per second. This performance demonstrates the effectiveness of the proposed enhancements in improving both accuracy and efficiency in defect detection.

Considering the defect detection issues in electroluminescence (EL) of photovoltaic (PV) cell systems, lots of factors result in performance degradation, including defect diversity, data imbalance, scale difference, etc. Focal-EIoU loss, an effective defect detection solution for EL, is proposed based on the improved YOLOv5. Firstly, by analyzing the detection background and scale characteristics of EL defects, a binary classification is carried out in the system. Subsequently, a cascade detection network based on YOLOv5 is designed to further extract features from the binary-classified defects. The defect localization and classification are achieved in this way. To address the problem of imbalanced defect samples, a loss function is designed based on EIoU and Focal-F1 Loss. Experimental results are illustrated to show the effectiveness. Compared with the existing CNN-based deep learning approaches, the proposed focal loss calculation-based method can effectively improve the performance of handling sample imbalance. Moreover, in the detection of 12 types of defects, the Yolov5 algorithms can always obtain higher MAP (mean average precision) even with different parameter levels (Yolov5m: 0.791 vs. 0.857, Yolov5l: 0.798 vs. 0.862, Yolov5x: 0.802 vs. 0.867, Yolov5s: 0.793 vs. 0.865).

To address the challenge of achieving a balance between efficiency and performance in steel surface defect detection, this paper presents a novel algorithm that enhances the YOLOv5 defect detection model. The enhancement process begins by employing the K-means ++ algorithm to fine-tune the location of the prior anchor boxes, improving the matching process. Subsequently, the loss function is transitioned from generalized intersection over union (GIOU) to efficient intersection over union (EIOU) to mitigate the former's degeneration issues. To minimize information loss, Carafe upsampling replaces traditional upsampling techniques. Lastly, the squeeze and excitation networks (SE-Net) module is incorporated to augment the model's sensitivity to channel features. Experimental evaluations conducted on a public defect dataset reveal that the proposed method elevates the mean average precision (mAP) by seven percentage points compared to the original YOLOv5 model, achieving an mAP of 83.3%. Furthermore, our model's size is significantly reduced compared to other advanced algorithms, while maintaining a processing speed of 47 frames per second. This performance demonstrates the effectiveness of the proposed enhancements in improving both accuracy and efficiency in defect detection. [ABSTRACT FROM AUTHOR]

This paper proposes an improved target detection algorithm, SDE-YOLO, based on the YOLOv5s framework, to address the low detection accuracy, misdetection, and leakage in blood cell detection caused by existing single-stage and two-stage detection algorithms. Initially, the Swin Transformer is integrated into the back-end of the backbone to extract the features in a better way. Then, the 32 × 32 network layer in the path-aggregation network (PANet) is removed to decrease the number of parameters in the network while increasing its accuracy in detecting small targets. Moreover, PANet substitutes traditional convolution with depth-separable convolution to accurately recognize small targets while maintaining a fast speed. Finally, replacing the complete intersection over union (CIOU) loss function with the Euclidean intersection over union (EIOU) loss function can help address the imbalance of positive and negative samples and speed up the convergence rate. The SDE-YOLO algorithm achieves a mAP of 99.5%, 95.3%, and 93.3% on the BCCD blood cell dataset for white blood cells, red blood cells, and platelets, respectively, which is an improvement over other single-stage and two-stage algorithms such as SSD, YOLOv4, and YOLOv5s. The experiment yields excellent results, and the algorithm detects blood cells very well. The SDE-YOLO algorithm also has advantages in accuracy and real-time blood cell detection performance compared to the YOLOv7 and YOLOv8 technologies. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Automated fruit-picking equipment has the potential to significantly enhance the efficiency of picking. Accurate detection and localization of fruits are particularly crucial in this regard. However, current methods rely on expensive tools such as depth cameras and LiDAR. This study proposes a low-cost method based on monocular images to achieve target detection and depth estimation. To improve the detection accuracy of targets, especially small targets, an advanced YOLOv8s detection algorithm is introduced. This approach utilizes the BiFormer block, an attention mechanism for dynamic query-aware sparsity, as the backbone feature extractor. It also adds a small-target-detection layer in the Neck and employs EIoU Loss as the loss function. Furthermore, a fused depth estimation method is proposed, which incorporates high-resolution, low-resolution, and local high-frequency depth estimation to obtain depth information with both high-frequency details and low-frequency structure. Finally, the spatial 3D coordinates of the fruit are obtained by fusing the planar coordinates and depth information. The experimental results with citrus as the target result in an improved YOLOv8s network mAP of 88.45% and a recognition accuracy of 94.7%. The recognition of citrus in a natural environment was improved by 2.7% compared to the original model. In the detection range of 30 cm~60 cm, the depth-estimation results (MAE, RSME) are 0.53 and 0.53. In the illumination intensity range of 1000 lx to 5000 lx, the average depth estimation results (MAE, RSME) are 0.49 and 0.64. In the simulated fruit-picking scenario, the success rates of grasping at 30 cm and 45 cm were 80.6% and 85.1%, respectively. The method has the advantage of high-resolution depth estimation without constraints of camera parameters and fruit size that monocular geometric and binocular localization do not have, providing a feasible and low-cost localization method for fruit automation equipment. [ABSTRACT FROM AUTHOR]

The modern tram track automatic cleaning car is a crucial equipment in urban rail transportation systems, effectively removing trash, dust, and other debris from the slotted tracks of trams. However, due to the complex and variable structure of turnouts, the cleaning car often requires assistance in accurately detecting their positions. Consequently, the cleaning car needs help in adequately cleaning or bypassing turnouts, which adversely affects cleaning effectiveness and track maintenance quality. This paper presents a novel method for tracking turnout identification called PBE-YOLO based on the improved yolov5s framework. The algorithm enhances yolov5s by optimizing the lightweight backbone network, improving feature fusion methods, and optimizing the regression loss function. The proposed method is trained using a dataset of track turnouts collected through field shots on modern tram lines. Comparative experiments are conducted to analyze the performance of the improved lightweight backbone network, as well as performance comparisons and ablation experiments for the new turnout identification method. Experimental results demonstrate that the proposed PBE-YOLO method achieves a 52.71% reduction in model parameters, a 4.60% increase in mAP@0.5(%), and a 3.27% improvement in precision compared to traditional yolov5s. By improving the track turnout identification method, this paper enables the automatic cleaning car to identify turnouts' positions accurately. This enhancement leads to several benefits, including increased automation levels, improved cleaning efficiency and quality, reduced reliance on manual intervention, and mitigation of collision risks between the cleaning car and turnouts. [ABSTRACT FROM AUTHOR]

Identifying the grade of Gastrodia elata in the market has low efficiency and accuracy. To address this issue, an I-YOLOX object detection algorithm based on deep learning and computer vision is proposed in this paper. First, six types of Gastrodia elata images of different grades in the Gastrodia elata planting cooperative were collected for image enhancement and labeling as the model training dataset. Second, to improve feature information extraction, an ECA attention mechanism module was inserted between the backbone network CSPDarknet and the neck enhancement feature extraction network FPN in the YOLOX model. Then, the impact of the attention mechanism and application position on model improvement was investigated. Third, the 3 × 3 convolution in the neck enhancement feature extraction network FPN and the head network was replaced by depthwise separable convolution (DS Conv) to reduce the model size and computation amount. Finally, the EIoU loss function was used to predict boundary frame regression at the output prediction end to improve the convergence speed of the model. The experimental results indicated that compared with the original YOLOX model, the mean average precision of the improved I-YOLOX network model was increased by 4.86% (97.83%), the model computation was reduced by 5.422 M (reaching 3.518 M), the model size was reduced by 20.6 MB (reaching 13.7 MB), and the image frames detected per second increased by 3 (reaching 69). Compared with other target detection algorithms, the improved model outperformed Faster R-CNN, SSD-VGG, YOLOv3s, YOLOv4s, YOLOv5s, and YOLOv7 algorithms in terms of mean average precision, model size, computation amount, and frames per second. The lightweight model improved the detection accuracy and speed of different grades of Gastrodia elata and provided a theoretical basis for the development of online identification systems of different grades of Gastrodia elata in practical production. [ABSTRACT FROM AUTHOR]

Target detection research has always been difficult when it comes to small target detection in underwater situations. To address the issues of a high miss detection rate and poor underwater scene recognition in underwater small target detection tasks, an improved underwater small target detection technique utilizing YOLOv7 is proposed. To achieve the accuracy rate while considering the high detection speed, the YOLOv7 network is used as the basic network. The network concentrates more crucial feature information of small targets to increase detection accuracy while reducing model complexity by merging the SENet attention mechanism, enhancing the FPN network topology, and incorporating the EIoU loss function. Through simulation tests, the mAP, P, and R metrics are confirmed on the test set and contrasted with other conventional target detection techniques. The outcomes demonstrate that the enhanced algorithm outperforms competing networks and successfully raises detection accuracy on the test set.

For the complex background along the electrified railroad, there are problems such as missed detection and false detection of invasive small target foreign objects. A high-precision recognition algorithm for small target foreign object detection is proposed. A channel attention mechanism ECA-Net is added to the backbone network of YOLOv5s so that the network can focus more on small target objects; borrowing from the network structure of weighted bidirectional feature pyramid network (BiFPN), the backbone network is fused with Neck by adding an edge to achieve cross-level feature fusion; the SPD-Conv module is used to replace the seventh layer convolutional kernel to increase the detail information extraction; using EIOU loss function to shorten the actual difference in width and height between the a priori frame and the real frame, The experimental results show that the improved algorithm increases the mAP value by 3.35 percentage points with little growth in computational volume, and the highest mAP value reaches 97.96%. The detection accuracy is greatly improved with a small loss of time cost, and the demand for small target foreign body detection in complex environments is achieved.

Copyright of Journal of Guangxi Normal University - Natural Science Edition is the property of Gai Kan Bian Wei Hui and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

With the development of wireless technology, signals propagating in space are easy to mix, so blind detection of communication signals has become a very practical and challenging problem. In this paper, we propose a blind detection method for broadband signals based on a weighted bi-directional feature pyramid network (BiFPN). The method can quickly perform detection and automatic modulation identification (AMC) on time-domain aliased signals in broadband data. Firstly, the method performs a time-frequency analysis on the received signals and extracts the normalized time-frequency images and the corresponding labels by short-time Fourier transform (STFT). Secondly, we build a target detection model based on YOLOv5 for time-domain mixed signals in broadband data and learn the features of the time-frequency distribution image dataset of broadband signals, which achieves the purpose of training the model. The main improvements of the algorithm are as follows: (1) a weighted bi-directional feature pyramid network is used to achieve a simple and fast multi-scale feature fusion approach to improve the detection probability; (2) the Efficient-Intersection over Union (EIOU) loss function is introduced to achieve high accuracy signal detection in a low Signal-Noise Ratio (SNR) environment. Finally, the time-frequency images are detected by an improved deep network model to complete the blind detection of time-domain mixed signals. The simulation results show that the method can effectively detect the continuous and burst signals in the broadband communication signal data and identify their modulation types. [ABSTRACT FROM AUTHOR]

Corona Virus Disease 2019 (COVID-19) poses a significant threat to human health and safety. As the core of the prevention and control of COVID-19, the health and safety of medical and nursing personnel are extremely important, and the standardized use of medical personal protective equipment can effectively prevent cross-infection. Due to the existence of severe occlusion and overlap, traditional image processing methods struggle to meet the demand for real-time detection. To address these problems, we propose the ME-YOLO model, which is an improved model based on the one-stage detector. To improve the feature extraction ability of the backbone network, we propose a feature fusion module (FFM) merged with the C3 module, named C3_FFM. To fully retain the semantic information and global features of the up-sampled feature map, we propose an up-sampling enhancement module (USEM). Furthermore, to achieve high-accuracy localization, we use EIoU as the loss function of the border regression. The experimental results demonstrate that ME-YOLO can better balance performance (97.2% mAP) and efficiency (53 FPS), meeting the requirements of real-time detection. [ABSTRACT FROM AUTHOR]

This paper proposes an improved target detection algorithm, SDE-YOLO, based on the YOLOv5s framework, to address the low detection accuracy, misdetection, and leakage in blood cell detection caused by existing single-stage and two-stage detection algorithms. Initially, the Swin Transformer is integrated into the back-end of the backbone to extract the features in a better way. Then, the 32 × 32 network layer in the path-aggregation network (PANet) is removed to decrease the number of parameters in the network while increasing its accuracy in detecting small targets. Moreover, PANet substitutes traditional convolution with depth-separable convolution to accurately recognize small targets while maintaining a fast speed. Finally, replacing the complete intersection over union (CIOU) loss function with the Euclidean intersection over union (EIOU) loss function can help address the imbalance of positive and negative samples and speed up the convergence rate. The SDE-YOLO algorithm achieves a mAP of 99.5%, 95.3%, and 93.3% on the BCCD blood cell dataset for white blood cells, red blood cells, and platelets, respectively, which is an improvement over other single-stage and two-stage algorithms such as SSD, YOLOv4, and YOLOv5s. The experiment yields excellent results, and the algorithm detects blood cells very well. The SDE-YOLO algorithm also has advantages in accuracy and real-time blood cell detection performance compared to the YOLOv7 and YOLOv8 technologies.

To overcome the problem of fewer sample and uneven distribution of defect type in defect detection of adhesive structure parts, a defect identification approach based on DCGAN and YOLOv5 is proposed. The above problems are solved by fine-tuning the structure and loss function of DCGAN, the generated high-quality defect images and the extended defect dataset are the basis for accurate identification with YOLOv5. The EIOU loss function is utilized in the YOLOv5 network, the mAP value and recall increase by 3.9% and 10.5% compared with the GIOU loss function, but the precision decreases. To solve this problem, the feature extraction capability of the network is enhanced by incorporating the CBAM after the C3 module in the YOLOv5 network. The mAP, precision, and recall of the optimized YOLOv5 algorithm are improved to 78.6%, 77.2%, and 76%, respectively, the precision compared to the original model improved by 10.6%. The results demonstrate that the improved YOLOv5 model can effectively identify defects of adhesive structure.

Fast, accurate, and non-destructive large-scale detection of sweet cherry ripeness is the key to determining the optimal harvesting period and accurate grading by ripeness. Due to the complexity and variability of the orchard environment and the multi-scale, obscured, and even overlapping fruit, there are still problems of low detection accuracy even using the mainstream algorithm YOLOX in the absence of a large amount of tagging data. In this paper, we proposed an improved YOLOX target detection algorithm to quickly and accurately detect sweet cherry ripeness categories in complex environments. Firstly, we took a total of 2400 high-resolution images of immature, semi-ripe, and ripe sweet cherries in an orchard in Hanyuan County, Sichuan Province, including complex environments such as sunny days, cloudy days, branch and leaf shading, fruit overlapping, distant views, and similar colors of green fruits and leaves, and formed a dataset dedicated to sweet cherry ripeness detection by manually labeling 36068 samples, named SweetCherry. On this basis, an improved YOLOX target detection algorithm YOLOX-EIoU-CBAM was proposed, which embedded the Convolutional Block Attention Module (CBAM) between the backbone and neck of the YOLOX model to improve the model's attention to different channels, spaces capability, and replaced the original bounding box loss function of the YOLOX model with Efficient IoU (EIoU) loss to make the regression of the prediction box more accurate. Finally, we validated the feasibility and reliability of the YOLOX-EIoU-CBAM network on the SweetCherry dataset. The experimental results showed that the method in this paper significantly outperforms the traditional Faster R-CNN and SSD300 algorithms in terms of mean Average Precision (mAP), recall, model size, and single-image inference time. Compared with the YOLOX model, the mAP of this method is improved by 4.12%, recall is improved by 4.6%, F-score is improved by 2.34%, while model size and single-image inference time remain basically comparable. The method in this paper can cope well with complex backgrounds such as fruit overlap, branch and leaf occlusion, and can provide a data base and technical reference for other similar target detection problems. [ABSTRACT FROM AUTHOR]

A lightweight rolled steel strip surface defect detection model, YOLOv5s-GCE, is proposed to improve the efficiency and accuracy of industrialized rolled steel strip defect detection. The Ghost module is used to replace the CBS structure in a part of the original YOLOv5s model, and the Ghost bottleneck is employed to replace the bottleneck structure in C3 to minimize the model's size and make the network lightweight. The EIoU function is added to improve the accuracy of the regression of the prediction frame and accelerate its convergence. The CA (Coordinate Attention) attention method is implemented to reinforce critical feature channels and their position information, enabling the model to identify and find targets correctly. The experimental results demonstrate that the accuracy of YOLOv5s-GCE is 85.7%, which is 3.5% higher than that of the original network; the model size is 7.6 MB, which is 44.9% smaller than that of the original network; the number of model parameters and calculations are reduced by 47.1% and 48.8%, respectively; and the detection speed reached 58.8 fps. YOLOv5s-GCE meets the necessity for real-time identification of rolled steel flaws in industrial production compared to other common algorithms. [ABSTRACT FROM AUTHOR]

Automated fruit-picking equipment has the potential to significantly enhance the efficiency of picking. Accurate detection and localization of fruits are particularly crucial in this regard. However, current methods rely on expensive tools such as depth cameras and LiDAR. This study proposes a low-cost method based on monocular images to achieve target detection and depth estimation. To improve the detection accuracy of targets, especially small targets, an advanced YOLOv8s detection algorithm is introduced. This approach utilizes the BiFormer block, an attention mechanism for dynamic query-aware sparsity, as the backbone feature extractor. It also adds a small-target-detection layer in the Neck and employs EIoU Loss as the loss function. Furthermore, a fused depth estimation method is proposed, which incorporates high-resolution, low-resolution, and local high-frequency depth estimation to obtain depth information with both high-frequency details and low-frequency structure. Finally, the spatial 3D coordinates of the fruit are obtained by fusing the planar coordinates and depth information. The experimental results with citrus as the target result in an improved YOLOv8s network mAP of 88.45% and a recognition accuracy of 94.7%. The recognition of citrus in a natural environment was improved by 2.7% compared to the original model. In the detection range of 30 cm~60 cm, the depth-estimation results (MAE, RSME) are 0.53 and 0.53. In the illumination intensity range of 1000 lx to 5000 lx, the average depth estimation results (MAE, RSME) are 0.49 and 0.64. In the simulated fruit-picking scenario, the success rates of grasping at 30 cm and 45 cm were 80.6% and 85.1%, respectively. The method has the advantage of high-resolution depth estimation without constraints of camera parameters and fruit size that monocular geometric and binocular localization do not have, providing a feasible and low-cost localization method for fruit automation equipment.

A fast and accurate detection is one of the most important prerequisites for the apple harvest robots. However, there are many factors that can make apple detection difficult in a real orchard scene, such as complex backgrounds, fruit overlap, and leaf/branch occlusion. In this study, a fast and stable network was proposed for apple detection using an improved RetinaNet. A picking strategy was also developed for the harvest robot. Specifically, once the apples occluded by branches/wires were regarded as the picking targets, the robot arm would be injured at the same time. Therefore, the apples were labeled with multiple classes, according to different types of occlusions. The Res2Net module was also embedded in the ResNet50, in order to improve the ability of the backbone network to extract the multi-scale features. Furthermore, the BiFPN instead of FPN was used as a feature fusion network in the neck of the network. A weight fusion of feature maps was also made at different scales for the apples with different sizes, thus improving the detection accuracy of the network. After that, a loss function was combined with the Focal loss and Efficient Intersection over Union (EIoU) loss. Among them, Focal loss was used for the classification loss function, further reducing the errors from the imbalance of positive and negative sample ratios. By contrast, the EIoU loss was used for the regression loss function of the bounding box, in order to maintain a fast and accurate regression. Particularly, there were some different relative positions in the prediction and the ground truth box, such as overlap, disjoint and inclusion. Finally, the classification and regression were carried out on the feature map of five scales to realize a better detection of apple. In addition, the original dataset consisted of 800 apple images with complex backgrounds of dense orchards. A data enhancement was conducted to promote the generalization ability of the model. The dataset was then expanded to 4 800 images after operations, such as rotating, adjusting brightness, and adding noise. There was also a balance between the detection accuracy and speed. A series of experimental statistics were obtained on the number of BiFPN stacks in the network. Specifically, the BiFPN was stacked five times in the improved RetinaNet. The ablation experiments showed that each improvement of the model enhanced the accuracy of the network for the apple detection, compared with the original. The average precision of the improved RetinaNet reached 94.02%, 86.74%, 89.42%, and 94.84% for the leaf occlusion, branch/wire occlusion, fruit occlusion, and no occlusion apples, respectively. The mean Average Precision (mAP) reached 91.26%, which was 5.02 percentage points higher than that of the traditional RetinaNet. The improved RetinaNet took only 42.72 ms to process an apple image on average. Correspondingly, each fruit picking cycle was 2 780 ms, indicating that the detection speed fully met the harsh requirement of the picking robot. Only when the apples were large or rarely occluded, both improved and traditional RetinaNet were used to accurately detect them. By contrast, the improved RetinaNet performed the best to detect all apple fruits, when the apples were under a complex environment in an orchard, such as the leaf-, fruit-, or branch/wire-occluded background. The reason was that the traditional RetinaNet often appeared to miss the detection in this case. Consequently, the best comprehensive performance was achieved to verify the effectiveness of the improvements, compared with the state-of-the-art detection network, such as the Faster RCNN and YOLOv4. Overall, all the apples in the different classes can be effectively detected for the apple harvest. The finding can greatly contribute to the picking strategy of the robot, further avoiding the potential damage by the branches and wires during harvesting. [ABSTRACT FROM AUTHOR]

Identifying the grade of Gastrodia elata in the market has low efficiency and accuracy. To address this issue, an I-YOLOX object detection algorithm based on deep learning and computer vision is proposed in this paper. First, six types of Gastrodia elata images of different grades in the Gastrodia elata planting cooperative were collected for image enhancement and labeling as the model training dataset. Second, to improve feature information extraction, an ECA attention mechanism module was inserted between the backbone network CSPDarknet and the neck enhancement feature extraction network FPN in the YOLOX model. Then, the impact of the attention mechanism and application position on model improvement was investigated. Third, the 3 × 3 convolution in the neck enhancement feature extraction network FPN and the head network was replaced by depthwise separable convolution (DS Conv) to reduce the model size and computation amount. Finally, the EIoU loss function was used to predict boundary frame regression at the output prediction end to improve the convergence speed of the model. The experimental results indicated that compared with the original YOLOX model, the mean average precision of the improved I-YOLOX network model was increased by 4.86% (97.83%), the model computation was reduced by 5.422 M (reaching 3.518 M), the model size was reduced by 20.6 MB (reaching 13.7 MB), and the image frames detected per second increased by 3 (reaching 69). Compared with other target detection algorithms, the improved model outperformed Faster R-CNN, SSD-VGG, YOLOv3s, YOLOv4s, YOLOv5s, and YOLOv7 algorithms in terms of mean average precision, model size, computation amount, and frames per second. The lightweight model improved the detection accuracy and speed of different grades of Gastrodia elata and provided a theoretical basis for the development of online identification systems of different grades of Gastrodia elata in practical production.

· A lightweight network enhances tile and bolt detection in CFETR. · Improving YOLOv5s with C3Ghost, SimSPPF, and EIOU structures. · GSE-YOLOv5s with higher mAP, fewer parameters, and less FLOPS. The detection of the first wall tile and bolts (holes) of CFETR is crucial for improving the efficiency of remote operation maintenance. The internal environment of CFETR is complex, and the existing target detection algorithms lack precision, are prone to missed detection, and occupy substantial computational resources. We introduce a method for detecting tiles and bolts (holes) using a lightweight network. First, a preliminary research platform is established to build a dataset. Secondly, based on YOLOv5s, the GhostBottleNeck module replaces the BottleNeck in the C3 module to build a C3Ghost structure, reducing model computation and parameters. A fast spatial pooling pyramid structure, SimSPPF module, is proposed to replace the SPPF module, increasing model speed. The replacement of the CIOU loss function with EIOU enhances positional accuracy. Comparative experimental results indicate this network outperforms baseline YOLOv5s by increasing the mAP by 1.4 %, decreasing parameters by 13.8 %, and reducing floating-point operations (FLOPS) by 13.2 %. This study holds considerable reference value for future remote visual maintenance in CFETR. [ABSTRACT FROM AUTHOR]

With the development of wireless technology, signals propagating in space are easy to mix, so blind detection of communication signals has become a very practical and challenging problem. In this paper, we propose a blind detection method for broadband signals based on a weighted bi-directional feature pyramid network (BiFPN). The method can quickly perform detection and automatic modulation identification (AMC) on time-domain aliased signals in broadband data. Firstly, the method performs a time-frequency analysis on the received signals and extracts the normalized time-frequency images and the corresponding labels by short-time Fourier transform (STFT). Secondly, we build a target detection model based on YOLOv5 for time-domain mixed signals in broadband data and learn the features of the time-frequency distribution image dataset of broadband signals, which achieves the purpose of training the model. The main improvements of the algorithm are as follows: (1) a weighted bi-directional feature pyramid network is used to achieve a simple and fast multi-scale feature fusion approach to improve the detection probability; (2) the Efficient-Intersection over Union (EIOU) loss function is introduced to achieve high accuracy signal detection in a low Signal-Noise Ratio (SNR) environment. Finally, the time-frequency images are detected by an improved deep network model to complete the blind detection of time-domain mixed signals. The simulation results show that the method can effectively detect the continuous and burst signals in the broadband communication signal data and identify their modulation types.

Corona Virus Disease 2019 (COVID-19) poses a significant threat to human health and safety. As the core of the prevention and control of COVID-19, the health and safety of medical and nursing personnel are extremely important, and the standardized use of medical personal protective equipment can effectively prevent cross-infection. Due to the existence of severe occlusion and overlap, traditional image processing methods struggle to meet the demand for real-time detection. To address these problems, we propose the ME-YOLO model, which is an improved model based on the one-stage detector. To improve the feature extraction ability of the backbone network, we propose a feature fusion module (FFM) merged with the C3 module, named C3_FFM. To fully retain the semantic information and global features of the up-sampled feature map, we propose an up-sampling enhancement module (USEM). Furthermore, to achieve high-accuracy localization, we use EIoU as the loss function of the border regression. The experimental results demonstrate that ME-YOLO can better balance performance (97.2% mAP) and efficiency (53 FPS), meeting the requirements of real-time detection.

A lightweight rolled steel strip surface defect detection model, YOLOv5s-GCE, is proposed to improve the efficiency and accuracy of industrialized rolled steel strip defect detection. The Ghost module is used to replace the CBS structure in a part of the original YOLOv5s model, and the Ghost bottleneck is employed to replace the bottleneck structure in C3 to minimize the model’s size and make the network lightweight. The EIoU function is added to improve the accuracy of the regression of the prediction frame and accelerate its convergence. The CA (Coordinate Attention) attention method is implemented to reinforce critical feature channels and their position information, enabling the model to identify and find targets correctly. The experimental results demonstrate that the accuracy of YOLOv5s-GCE is 85.7%, which is 3.5% higher than that of the original network; the model size is 7.6 MB, which is 44.9% smaller than that of the original network; the number of model parameters and calculations are reduced by 47.1% and 48.8%, respectively; and the detection speed reached 58.8 fps. YOLOv5s-GCE meets the necessity for real-time identification of rolled steel flaws in industrial production compared to other common algorithms.

To accurately recognize plant diseases under complex natural conditions, an improved plant disease-recognition model based on the original YOLOv5 network model was established. First, a new InvolutionBottleneck module was used to reduce the numbers of parameters and calculations, and to capture long-distance information in the space. Second, an SE module was added to improve the sensitivity of the model to channel features. Finally, the loss function ‘Generalized Intersection over Union’ was changed to ‘Efficient Intersection over Union’ to address the former’s degeneration into ‘Intersection over Union’. These proposed methods were used to improve the target recognition effect of the network model. In the experimental phase, to verify the effectiveness of the model, sample images were randomly selected from the constructed rubber tree disease database to form training and test sets. The test results showed that the mean average precision of the improved YOLOv5 network reached 70%, which is 5.4% higher than that of the original YOLOv5 network. The precision values of this model for powdery mildew and anthracnose detection were 86.5% and 86.8%, respectively. The overall detection performance of the improved YOLOv5 network was significantly better compared with those of the original YOLOv5 and the YOLOX_nano network models. The improved model accurately identified plant diseases under natural conditions, and it provides a technical reference for the prevention and control of plant diseases.

Cell micronucleus detection is an significant method used to evaluate damage to human cells by drugs, radiation, toxic substances, etc. Aiming at the problems of small target size, high similarity between classes and small sample data in cell micronucleus image recognition, this paper proposes an improved YOLOv5 cell micronucleus high-through detection algorithm named YOLOv5-CEB. Firstly, the basic network is optimized by replacing loss function, resetting anchor box and performing exploratory experiments on network depth. Then, in order to enhance the feature extraction ability of the network, the Bifpn structure is introduced. After that, the semantic connection of context is enhanced by the mechanism of fusing global and local attention, which makes the network pays more attention to ROI. Finally,a shallow detection layer is added, which improve the accuracy of cell micronucleus recognition by multi-scale detection. Compared with the initial YOLOv5 algorithm, the experimental results show that the average precision, accuracy and recall of the YOLOv5-CEB algorithm are improved by 2%, 2% and 1.3%, respectively. [ABSTRACT FROM AUTHOR]