Search

In maritime surveillance systems, object detection plays a crucial role in ensuring the security of nearby waters by tracking the movement of various objects, such as ships and aircrafts, that are found at sea, detecting illegal activities and preemptively countering or predicting potential risks. Using vision sensors such as cameras to monitor the sea can help to identify the shape, size, and color of objects, enabling the precise analysis of maritime situations. Additionally, vision sensors can monitor or track small ships that may escape radar detection. However, objects located at considerable distances from vision sensors have low resolution and are small in size, rendering their detection difficult. This paper proposes a multifocal object detection associative network (MODAN) to overcome these vulnerabilities and provide stable maritime surveillance. First, it searches for the horizon using color quantization based on K-means; then, it selects and partitions the region of interest (ROI) around the horizon using the ROI selector. The original image and ROI image, converted to high resolution through the Super-Resolution Convolutional Neural Network (SRCNN), are then passed to the near-field and far-field detectors, respectively, for object detection. The weighted box fusion removes duplicate detected objects and estimates the optimal object. The proposed network is more stable and efficient in detecting distant objects than existing single-object detection models. Through performance evaluations, the proposed network exhibited an average precision surpassing that of the existing single-object detection models by more than 7%, and the false detection rate was reduced by 59% compared to similar multifocal-based state-of-the-art detection methods. [ABSTRACT FROM AUTHOR]

The rapid progress in deep learning technologies has accelerated the use of object detection models, but most models do not operate satisfactorily in low-light environments. As a result, many studies have been conducted on image enhancement techniques aiming to make objects more visible by increasing contrast, but the process of image enhancement may negatively impact detection as it further strengthens unwanted noises due to indirect factors of light reflection such as overall low brightness, streetlamps, and neon signboards. Therefore, in this study, we propose a technique for improving the performance of object detection in low-light environments. The proposed technique inverts a low-light image to make it similar to a hazy image and then uses a haze removal algorithm based on entropy and fidelity to increase image contrast, clarifying the boundary between the object and the background. In the next step, we used the adaptive 2D Wiener filter (A2WF) to attenuate the noise accidentally strengthened during the image enhancement process and reinforced the boundary between the object and the background to increase detection performance. The test evaluation results showed that the proposed image enhancement scheme significantly increased image perception performance with the perception-based image quality evaluator being 12.73% lower than existing image enhancement techniques. In a comparison of vehicle detection performance, the proposed technique for enhancing nighttime images combined with the detection model proved its effectiveness by increasing the average precision by up to 18.63% against existing detection methods. [ABSTRACT FROM AUTHOR]

Recently, the rapid development of convolutional neural networks (CNN) has consistently improved object detection performance using CNN and has naturally been implemented in autonomous driving due to its operational potential in real-time. Detecting moving targets to realize autonomous driving is an essential task for the safety of drivers and pedestrians, and CNN-based moving target detectors have shown stable performance in fair weather. However, there is a considerable drop in detection performance during poor weather conditions like hazy or foggy situations due to particles in the atmosphere. To ensure stable moving object detection, an image restoration process with haze removal must be accompanied. Therefore, this paper proposes an image dehazing network that estimates the current weather conditions and removes haze using the haze level to improve the detection performance under poor weather conditions due to haze and low visibility. Combined with the thermal image, the restored image is assigned to the two You Only Look Once (YOLO) object detectors, respectively, which detect moving targets independently and improve object detection performance using late fusion. The proposed model showed improved dehazing performance compared with the existing image dehazing models and has proved that images taken under foggy conditions, the poorest weather for autonomous driving, can be restored to normal images. Through the fusion of the RGB image restored by the proposed image dehazing network with thermal images, the proposed model improved the detection accuracy by up to 22% or above in a dense haze environment like fog compared with models using existing image dehazing techniques. [ABSTRACT FROM AUTHOR]

For autonomous vehicles, it is critical to be aware of the driving environment to avoid collisions and drive safely. The recent evolution of convolutional neural networks has contributed significantly to accelerating the development of object detection techniques that enable autonomous vehicles to handle rapid changes in various driving environments. However, collisions in an autonomous driving environment can still occur due to undetected obstacles and various perception problems, particularly occlusion. Thus, we propose a robust object detection algorithm for environments in which objects are truncated or occluded by employing RGB image and light detection and ranging (LiDAR) bird's eye view (BEV) representations. This structure combines independent detection results obtained in parallel through "you only look once" networks using an RGB image and a height map converted from the BEV representations of LiDAR's point cloud data (PCD). The region proposal of an object is determined via non-maximum suppression, which suppresses the bounding boxes of adjacent regions. A performance evaluation of the proposed scheme was performed using the KITTI vision benchmark suite dataset. The results demonstrate the detection accuracy in the case of integration of PCD BEV representations is superior to when only an RGB camera is used. In addition, robustness is improved by significantly enhancing detection accuracy even when the target objects are partially occluded when viewed from the front, which demonstrates that the proposed algorithm outperforms the conventional RGB-based model. [ABSTRACT FROM AUTHOR]

The field of research related to video data has difficulty in extracting not only spatial but also temporal features and human action recognition (HAR) is a representative field of research that applies convolutional neural network (CNN) to video data. The performance for action recognition has improved, but owing to the complexity of the model, some still limitations to operation in real-time persist. Therefore, a lightweight CNN-based single-stream HAR model that can operate in real-time is proposed. The proposed model extracts spatial feature maps by applying CNN to the images that develop the video and uses the frame change rate of sequential images as time information. Spatial feature maps are weighted-averaged by frame change, transformed into spatiotemporal features, and input into multilayer perceptrons, which have a relatively lower complexity than other HAR models; thus, our method has high utility in a single embedded system connected to CCTV. The results of evaluating action recognition accuracy and data processing speed through challenging action recognition benchmark UCF-101 showed higher action recognition accuracy than the HAR model using long short-term memory with a small amount of video frames and confirmed the real-time operational possibility through fast data processing speed. In addition, the performance of the proposed weighted mean-based HAR model was verified by testing it in Jetson NANO to confirm the possibility of using it in low-cost GPU-based embedded systems. [ABSTRACT FROM AUTHOR]

An essential component for the autonomous flight or air-to-ground surveillance of a UAV is an object detection device. It must possess a high detection accuracy and requires real-time data processing to be employed for various tasks such as search and rescue, object tracking and disaster analysis. With the recent advancements in multimodal data-based object detection architectures, autonomous driving technology has significantly improved, and the latest algorithm has achieved an average precision of up to 96%. However, these remarkable advances may be unsuitable for the image processing of UAV aerial data directly onboard for object detection because of the following major problems: (1) Objects in aerial views generally have a smaller size than in an image and they are uneven and sparsely distributed throughout an image; (2) Objects are exposed to various environmental changes, such as occlusion and background interference; and (3) The payload weight of a UAV is limited. Thus, we propose employing a new real-time onboard object detection architecture, an RGB aerial image and a point cloud data (PCD) depth map image network (RGDiNet). A faster region-based convolutional neural network was used as the baseline detection network and an RGD, an integration of the RGB aerial image and the depth map reconstructed by the light detection and ranging PCD, was utilized as an input for computational efficiency. Performance tests and evaluation of the proposed RGDiNet were conducted under various operating conditions using hand-labeled aerial datasets. Consequently, it was shown that the proposed method has a superior performance for the detection of vehicles and pedestrians than conventional vision-based methods. [ABSTRACT FROM AUTHOR]