Search

The identification of plant nematodes is crucial in the fields of pest control, soil ecology, and biogeography. The automated recognition of plant nematodes based on deep-learning technology can significantly improve the accuracy and efficiency of their recognition. In this study, we devised a method for the multi-mode, multi-feature identification of plant nematodes using deep-learning techniques which emulated the recognition logic of domain experts. Beginning with a multi-featured plant nematode dataset, we not only designed key feature extraction strategies to address the problem of weak key feature points and small inter-specific differences in plant nematodes but also proposed a multi-feature joint training scheme and constructed a neural network structure with interpretability. Finally, an intelligent decision-making expert identification system for plant nematodes was implemented, and its performance was tested on the multi-feature plant nematode dataset. The results indicate that our model achieves an accuracy of up to 96.74% in identifying 23 species across two-body parts, which is 17.5% higher than the single-part feature identification. The accuracy of identifying 11 species in three-body parts reached 98.46%, an improvement of 1.24% over that of the two-part feature identification. Our novel model demonstrates that the accuracy of the expert system can be increased by incorporating more nematode feature parts.

Surface defect detection is a key task in industrial production processes. However, the existing methods still suffer from low detective accuracy to pit and small defects. To solve those problems, we establish a dataset of pit defects and propose a prior model-guided network for defect detection. This network is composed of a segmentation network with a weight label, classification network, and pyramid feature fusion module. The segmentation network as the prior model can improve the accuracy of the classification network. The weight label with center distance transformation can reduce the label cost of the segmentation network. The pyramid feature fusion module can adapt defects of different scales and avoid information loss of small defects. The comparison experiments are implemented to identify the performance of the proposed network. Ablation experiments are designed to specify the effectiveness of every module. Finally, the network is performed on a public dataset to verify its robustness. Experimental results reveal that the proposed method can effectively identify pit defects of different scales and improve the accuracy of defect detection. The accuracy can reach 99.3%, which is increased by 2~5% compared with other methods, revealing its excellent applicability in automatic quality inspection of industrial production.

Neural radiation field (NeRF)-based novel view synthesis methods are gaining popularity. NeRF can generate more detailed and realistic images than traditional methods. Conventional NeRF reconstruction of a room scene requires at least several hundred images as input data and generates several spatial sampling points, placing a tremendous burden on the training and prediction process with respect to memory and computational time. To address these problems, we propose a prior-driven NeRF model that only accepts sparse views as input data and reduces a significant number of non-functional sampling points to improve training and prediction efficiency and achieve fast high-quality rendering. First, this study uses depth priors to guide sampling, and only a few sampling points near the controllable range of the depth prior are used as input data, which reduces the memory occupation and improves the efficiency of training and prediction. Second, this study encodes depth priors as distance weights into the model and guides the model to quickly fit the object surface. Finally, a novel approach combining the traditional mesh rendering method (TMRM) and the NeRF volume rendering method was used to further improve the rendering efficiency. Experimental results demonstrated that our method had significant advantages in the case of sparse input views (11 per room) and few sampling points (8 points per ray).

Aims: Carbon monoxide poisoning is a common condition that can cause severe neurological sequelae. Previous studies have revealed that functional connectivity in carbon monoxide poisoning is abnormal under the assumption that it is resting during scanning and have focused on studying delayed encephalopathy in carbon monoxide poisoning. However, studies of functional connectivity dynamics in the acute phase of carbon monoxide poisoning may provide a more insightful perspective for understanding the neural mechanisms underlying carbon monoxide poisoning. To our knowledge, this is the first study that explores abnormal brain network dynamics in the acute phase of carbon monoxide poisoning.Methods: Combining the sliding window method and k-means algorithm, we identified four recurrent dynamic functional cognitive impairment states from resting-state functional magnetic resonance imaging data from 29 patients in the acute phase of carbon monoxide poisoning and 29 healthy controls. We calculated between-group differences in the temporal properties and intensity of dFC states, and we also performed subgroup analyses to separately explore the brain network dynamics characteristics of adult vs. child carbon monoxide poisoning groups. Finally, these differences were correlated with patients’ cognitive performance in the acute phase of carbon monoxide poisoning and coma duration.Results: We identified four morphological patterns of brain functional network connectivity. During the acute phase of carbon monoxide poisoning, patients spent more time in State 2, which is characterized by positive correlation between SMN and CEN, and negative correlation between DMN and SMN. In addition, the fractional window and mean dwell time of State 2 were positively correlated with coma duration. The subgroup analysis results demonstrated that the acute phase of childhood carbon monoxide poisoning had greater dFNC time variability than adult carbon monoxide poisoning.Conclusion: Our findings reveal that patients in the acute phase of carbon monoxide poisoning exhibit dynamic functional abnormalities. Furthermore, children have greater dFNC instability following carbon monoxide poisoning than adults. This advances our understanding of the pathophysiological mechanisms underlying acute carbon monoxide poisoning.

High-precision, high-speed detection and classification of weakly differentiated targets has always been a difficult problem in the field of image vision. In this paper, the detection of phytopathogenic Bursaphelenchus xylophilus with small size and very weak inter-species differences is taken as an example. Our work is aimed at the current problem of weakly differentiated target detection: We propose a lightweight self attention network. Experiments show that the key feature recognition areas of plant nematodes found by our Self Attention network are in good agreement with the experience and knowledge of customs experts, and the feature areas found by this method can obtain higher detection accuracy than expert knowledge; In order to optimize the computing power brought by the whole image input, we use low resolution images to quickly obtain the location coordinates of key features, and then obtain the information of high resolution feature regions based on the coordinates; The adaptive weighted multi feature joint detection method based on heat map brightness is adopted to further improve the detection accuracy; We have constructed a more complete high-resolution training data set, involving 24 species of Equisetum and other common hybrids, with a total data volume of more than 10,000. The algorithm proposed in this paper replaces the tedious extensive manual labelling in the training process, improves the average training time of the model by more than 50%, reduces the testing time of a single sample by about 27%, optimizes the model storage size by 65%, improves the detection accuracy of the ImageNet pre-trained model by 12.6%, and improves the detection accuracy of the no-ImageNet pre-trained model by more than 48%.

Background: High-precision, high-speed detection and classification of weakly differentiated targets has always been a difficult problem in the field of image vision. In this paper, the detection of phytopathogenic Bursaphelenchus xylophilus with small size and very weak inter-species differences is taken as an example. Methods: Our work has been carried out in response to the current weakly differentiated target detection problems: a. To replace the complex network labelling and training process based on expert empirical knowledge, we proposed a lightweight Self-Attention network. Experiments proved that the key feature identification areas of​ plant nematodes found by our Self-Attention network is in good agreement with the customs expert empirical knowledge, and the feature areas found by the method in this paper can better obtain higher detection accuracy than the expert knowledge; b. To optimize the computing power caused by the input of the entire image, we used low-resolution images to quickly obtain the key feature location coordinates, then obtain high-resolution feature areas information based on the coordinates; c. We adopt the adaptive weight with multi-feature joint detection method based on the brightness of the heatmap to further improve the detection accuracy; d. We constructed a more complete high-resolution training dataset involving 24 species of Bursaphelenchus xylophilus and other common hybrid species with a total amount of data exceeded 10,000. Results: The algorithm proposed in this paper replaces the tedious extensive manual labelling in the training process, improves the average training time of the model by more than 50%, reduces the testing time of a single sample by about 27%, optimizes the model storage size by 65%, improves the detection accuracy of the ImageNet pre-trained model by 12.6%, and improves the detection accuracy of the no-ImageNet pre-trained model by more than 48%. Conclusions: Overall, the method in this paper has achieved better results in terms of model complexity, training annotation, as well as testing time and accuracy.

Bursaphelenchus is the key quarantine object of the customs quarantine department, but there are many kinds of Bursaphelenchus and their morphological changes are very complicated. Therefore, the identification of Bursaphelenchus has long been a difficult problem in the quarantine work of various customs. We propose a new identification method based on convolutional neural network with pluggable attention module. The method first uses the convolutional neural network to extract the features of Bursaphelenchus, and then obtains the attention map of the characteristic part of Bursaphelenchus through the pluggable attention module, adds the attention map to the identification network for training, and then strengthens the attention of different region weights, thereby strengthening the neural network to focus on learning the key feature regions of nematodes, so that the network can more accurately identify effective feature information. Compared with the traditional neural network model, the algorithm model proposed in this paper has achieved a high recognition accuracy rate and can effectively identify samples that are difficult to distinguish. The algorithm module proposed in this paper has been able to applied to the identification and classification of Bursaphelenchus in the quarantine department of Ningbo Customs.

The demands of the application of deep learning for the detection and marking of fine-grained object in the large field of view are increasingly prominent, which can be seen from automatic driving, traffic sign detection, satellite image analysis and so on. Most of the current studies focusing on the fine-grained object detection make an improvement based on the existing object detection algorithms to increase the detection accuracy of fine-grained object. This paper will propose a novel algorithm based on neural network feature constraints, which can realize the detection and marking of fine-grained object via network with the guidance of an attention map. In the procedures of neural network training, the Attention Mask is employed to constrain the loss function of the network and extract feature maps of key areas to alter the weights of key features through self-adaption. In this paper, combined with the needs of nematode detection project, taking nematode detection as an example, the ablation experiments with the employment of UNet network demonstrate that the accuracy rate of fine-grained object detection is increased from zero to around 85% with the additional loss function constraints.

A novel type of inorganic polymer coagulant, polyaluminum zirconium silicate sulfate (PAZS), was synthesized via co-polymerization. In the treatment of kaolin-humic acid synthetic water, it was found to be most effective when Si/(Zr + Al) = 1, OH/(Al + Zr) = 0.1, and Al/Zr = 3. Through CCD design, zeta potential and floc analysis, PAZS showed excellent coagulation and sedimentation performance in an experimental pH range of 3∼8. The prepared composite coagulant PAZS was characterized by FTIR, XRD, and SEM. Results indicate that PAZS is a highly polymerized complex with rough folded surfaces. The adsorption bridging effect and charge characteristics enabled PAZS to function in an acidic environment. The findings from the research conducted indicate that PAZS is a promising coagulant for efficient water treatment.

In this paper, the thermal energy storage system of Badaling 1 MW solar power tower plant is modelled from mathematical models for whole of the working conditions using the modular modelling method. This model can accurately simulate the recharge and discharge processes of thermal energy storage system. The dynamic and static characteristics of the thermal energy storage system are analyzed based on the model response curves of the system state parameters that are obtained from different steam flow disturbances. Conclusions of this paper are good references for the design, operating, and control strategy of solar thermal power plant.

Dimensionality reduction is a frequent preprocessing step in hyperspectral image analysis. High-dimensional data will cause the issue of the “curse of dimensionality” in the applications of hyperspectral imagery. In this paper, a dimensionality reduction method of hyperspectral images based on random projection (RP) for target detection was investigated. In the application areas of hyperspectral imagery, e.g. target detection, the high dimensionality of the hyperspectral data would lead to burdensome computations. Random projection is attractive in this area because it is data independent and computationally more efficient than other widely-used hyperspectral dimensionality-reduction methods, such as Principal Component Analysis (PCA) or the maximum-noise-fraction (MNF) transform. In RP, the original highdimensional data is projected onto a low dimensional subspace using a random matrix, which is very simple. Theoretical and experimental results indicated that random projections preserved the structure of the original high-dimensional data quite well without introducing significant distortion. In the experiments, Constrained Energy Minimization (CEM) was adopted as the target detector and a RP-based CEM method for hyperspectral target detection was implemented to reveal that random projections might be a good alternative as a dimensionality reduction tool of hyperspectral images to yield improved target detection with higher detection accuracy and lower computation time than other methods.

A sparse representation based multi-threshold segmentation (SRMTS) algorithm for target detection in hyperspectral images is proposed. Benefiting from the sparse representation, the high-dimensional spectral data can be characterized into a series of sparse feature vectors which has only a few nonzero coefficients. Through setting an appropriate threshold, the noise removed sparse spectral vectors are divided into two subspaces in the sparse domain consistent with the sample spectrum to separate the target from the background. Then a correlation and a vector 1-norm are calculated respectively in the subspaces. The sparse characteristic of the target is used to ext ract the target with a multi -threshold method. Unlike the conventional hyperspectral dimensionality reduction methods used in target detection algorithms, like Principal Components Analysis (PCA) and Maximum Noise Fraction (MNF), this algorithm maintains the spectral characteristics while removing the noise due to the sparse representation. In the experiments, an orthogonal wavelet sparse base is used to sparse the spectral information and a best contraction threshold to remove the hyperspectral image noise according to the noise estimation of the test images. Compared with co mmon algorithms, such as Adaptive Cosine Estimator (ACE), Constrained Energy Minimizat ion (CEM) and the noise removed MNF-CEM algorithm, the proposed algorithm demonstrates higher detection rates and robustness via the ROC curves.

In imaging systems based on compressed sensing, error in the measured data is incurred due to the nonlinear response of the photo-detector, which affects the quality of reconstructed images. Conventionally, the affected measured data will be eliminated in order to obtain the high reconstruction quality. However, when there are too many measured data affected, it is impossible to reject all such data which will certainly degrade the imaging efficiency. Therefore, an algorithm of regionally compensating the non-linear response from the detector is proposed. The nonlinear measured data will be compensated but not rejected in the proposed algorithm. According to the detector response curve, all of the nonlinear measured data will be divided into several parts. The data in the same part will have the same average compensation factor which can be obtained from the response curve after calculation. The affected measured data will be compensated with the compensation factor regionally before used to reconstruct the original image. The theoretical analysis and simulation results show that a reconstructed image with high quality can still be obtained even when over 80% of the measured data are nonlinear. It is impossible to get such a high quality reconstructed image with the conventionally algorithm. The PSNR and the M-rate in the simulation show that the compensation algorithm can greatly deal with the situation of too many nonlinear measured data.

Based on the principles of mass, energy conservation and state equation for ideal gas, operation performance of two-stage thermal energy storage system of the first 1MW solar tower power plant in China was calculated and analyzed using numerical simulation method. The results indicated that the pressure of nitrogen in oil tanks slightly changes as a function of temperature and volume variation, thus it is not necessary to control pressure by nitrogen recharge and freeing in normal operation. The steam mass flow rate of discharging process is dependent on pressure variation rate of steam accumulator. Steam mass flow rate increases rapidly to the controlled maximum while pressure varies with time as exponential function. For the case of pressure decreases with time as linear function, the steam mass flow rate increases with time. The steam accumulator needs periodic water supplement because a small quantity of water was lost after every cycle of charging and discharging process.

The theory of compressed sensing (CS) indicates that a signal that is sparse or compressible can be recovered from a relatively small number of nonadaptive linear measurements that is far below the Nyquist-Shannon limit. However, CS suffers from a huge stored and computational overhead when dealing with images of high resolution, taking tens of minutes or longer. In this work, we extend the concept of wavelet trees by adding the sibling relationship and propose an imaging strategy named adaptive compressed sampling based on extended wavelet trees (EWT-ACS). Exploiting both parent-children relationship and sibling relationship in extended wavelet trees, EWT-ACS predicts the locations of significant coefficients adaptively and samples the significant coefficients using a binary digital micromirror device directly. The simulation and experimental results reveal that the proposed strategy breaks through the limitation in CS, and the reconstruction time is reduced significantly. Due to its single-pixel detection mechanism, EWT-ACS shows great potential in many imaging applications.

In imaging systems based on compressed sensing, error in the measured data is incurred due to the nonlinear response of the photo detector, which affects the quality of the reconstructed images. We propose an algorithm to compensate the nonlinear response from the detector. The compensation is based on the detector response curve on the measured data. The theoretical analysis and simulation results show that this algorithm can greatly reduce the reconstruction errors caused by the detector’s nonlinear response. Furthermore the peak signal-to-noise ratio of the reconstructed image and the system reconstruction rate have been significantly improved, while the fine feature of the images is better preserved and reconstructed as compared to that without using the algorithm.