Accurate detection of independent moving objects is the key to ensure the safety of driverless vehicles during travelling. In this study, a moving object detection method based on a vehicle-mounted binocular camera is proposed. Firstly, a feature matching points selection strategy is designed for high accuracy camera ego-motion estimation. Then, the residual optical flow filed generated only by moving objects is estimated by taking into account the camera ego-motion and global mixed optical flow. A dynamic threshold segmentation method based on disparity is proposed to separate the moving regions from the residual optical flow field. Finally, 3D and 2D information are combined for extracting each single moving object from moving regions. The innovation of the proposed method is that it can detect any type and any size of moving object theoretically by using the dense motion information of the images. By taking the data in KITTI database as the test samples, the detection precision of the moving objects using the proposed algorithm can reach up to 91% without considering tracking strategy.

Image edge detection is a fundamental problem in image processing and computer vision, particularly in the area of feature extraction. However, the time complexity increases squarely with the increase of image resolution in conventional serial computing mode. This results in being unbearably time consuming when dealing with a large amount of image data. In this paper, a novel resolution free parallel implementation algorithm for gradient based edge detection, namely EDENP, is proposed. The key point of our method is the introduction of an enzymatic numerical P system (ENPS) to design the parallel computing algorithm for image processing for the first time. The proposed algorithm is based on a cell-like P system with a nested membrane structure containing four membranes. The start and stop of the system is controlled by the variables in the skin membrane. The calculation of edge detection is performed in the inner three membranes in a parallel way. The performance and efficiency of this algorithm are evaluated on the CUDA platform. The main advantage of EDENP is that the time complexity of $O\left(1\right)$ can be achieved regardless of image resolution theoretically.

In this paper, a tracking algorithm based on the residual neural network model and machine learning is proposed. Compared with the widely used VGG network, the residual neural network has deeper characteristic layers and special additional layer structure, which break the symmetry of the network and reduce the degradation of the neural network. The additional layer and convolution layer are used for feature fusion to represent the target. The multi-features of the object can be captured by using the developed algorithm, so that the accuracy of tracking can be improved in some complex scenarios. In addition, we defined a new measure to calculate the similarity of different image regions and find the optimal matched region. The search area is delimited according to the continuity of the target motion, which improves the real-time performance of tracking. The experimental results illustrate that the proposed algorithm achieved a higher accuracy while taking into account the real time performance, especially in dealing with some complex scenarios such as deformation, rotation changes and background clutters, in comparison with the Multi-Domain Network (MDNet) algorithm based on a convolutional neural network.

Silk cocoons are an important raw material for textiles. The mixture of defective cocoons, such as thin-shelled cocoons, double-pupa cocoons, and yellow-spotted cocoons, can lead to issues during the reeling process, such as thread breakage, uneven material, or contamination, which diminishes the quality of the silk. This paper addresses the cocoon sorting problem by leveraging the image characteristics of defective cocoons and proposes a CD-YOLO defect classification and recognition method based on YOLOv7. First, an image acquisition system based on a line array camera is established to achieve efficient and high-quality cocoon image collection. Data augmentation is applied to the sample images, and a cocoon dataset is created. Then, using the structural characteristics of a grid-clustered layout, the original cocoon images are pre-segmented, and the K-Canopy algorithm is used for clustering to accurately obtain anchor boxes that match the cocoon size and complete the segmentation of individual cocoon images. Next, model improvements are made by adding a CBAM module to the backbone network, incorporating the SE attention mechanism into the neck network's convolution, and replacing the detection head with the Decoupled_Detect head to establish the CD-YOLO network model. Finally, the algorithm is tested and compared on the developed intelligent cocoon selection system. The experimental results show that the grid-clustered cocoon quality detection algorithm based on CD-YOLO outperforms the original network model, with the mean average precision (mAP) increasing by 5.8%. The mAP for defects at 0.5 threshold reaches 0.975, and the detection speed improves by 60%, achieving 45 fps. The optimized algorithm achieves a cocoon classification accuracy of ≥97% and meets the demand for a cocoon picking rate of ≥300kg per hour. This significantly enhances the sorting efficiency and accuracy, providing technological support for the automation upgrade in the sericulture industry.

Autonomous intelligence plays a significant role in aviation security. Since most aviation accidents occur in the take-off and landing stage, accurate tracking of moving object in airport apron will be a vital approach to ensure the operation of the aircraft safely. In this study, an adaptive object tracking method based on a discriminant is proposed in multi-camera panorama surveillance of large-scale airport apron. Firstly, based on channels of color histogram, the pre-estimated object probability map is employed to reduce searching computation, and the optimization of the disturbance suppression options can make good resistance to similar areas around the object. Then the object score of probability map is obtained by the sliding window, and the candidate window with the highest probability map score is selected as the new object center. Thirdly, according to the new object location, the probability map is updated, the scale estimation function is adjusted to the size of real object. From qualitative and quantitative analysis, the comparison experiments are verified in representative video sequences, and our approach outperforms typical methods, such as distraction-aware online tracking, mean shift, variance ratio, and adaptive colour attributes.