Abstract
The preprocessed images are input to a pretrained neural network to obtain the corresponding feature mapping, and the corresponding region of interest is set for each point in the feature mapping to obtain multiple candidate feature regions; subsequently, these candidate feature regions are fed into a region proposal network and a deep residual network for binary classification and BB regression, and some of the candidate feature regions are filtered out, and the remaining feature regions are subjected to ROIAIign operation; finally, classification, BB regression, and mask generation are performed on these feature regions, and full convolutional nerve network operation is performed in each feature region and output. To further identify the specific model of the vehicle, this paper proposes a multifeature model recognition method that fuses the improved model with the optimized Mask R-CNN algorithm. A vehicle local feature dataset including vehicle badges, lights, air intake grille, and whole vehicle outline is established to simplify the network structure of model. Meanwhile, its detection frame generation process and the adjustment rules of overlapping frame confidence in nonmaximum suppression are improved for coarse vehicle localization. Then, the generated vehicle detection frames after localization are output to the Mask R-CNN algorithm after further optimizing the RPN structure. The localized vehicle detection frames are then output to the Mask R-CNN algorithm after further optimization of the RPN structure for local feature recognition, and good recognition results are achieved. Finally, this paper establishes a distributed server-based vehicle recognition system, which mainly includes database module, file module, feature extraction and matching module, message queue module, WEB module, and vehicle detection module. Due to the limitations of traditional region generation methods, this paper provides a brief analysis of the region generation network in the Faster R-CNN algorithm and details the loss calculation principle of the output layer.
Highlights
The number of motor vehicles has exceeded 350 million, cars reached 229 million, motor vehicle drivers exceeded 420 million, including 360 million car drivers, and cars have gradually replaced bicycles and other as one of the main means of transportation for travel, appearing in various scenes such as streets, highways, and communities [1]
This paper mainly focuses on deep learning and convolutional neural network algorithms to optimize the network structure to train the detection and recognition models of large class vehicles and fine vehicles, respectively
Based on the algorithm development of R-CNN and Faster R-CNN and the design of convolutional layer, the superiority of convolutional neural network in target detection and recognition is illustrated, and the advantages and disadvantages of different methods and network frameworks in target detection are analyzed, and the improved Mask R-CNN method is proposed to recognize large classes of vehicles, and the components and functions of the improved algorithm are introduced in detail
Summary
The number of motor vehicles has exceeded 350 million, cars reached 229 million, motor vehicle drivers exceeded 420 million, including 360 million car drivers, and cars have gradually replaced bicycles and other as one of the main means of transportation for travel, appearing in various scenes such as streets, highways, and communities [1]. The core of the vehicle detection and automatic identification system construction lies in the license plate, vehicle color, vehicle brand, and specific model recognition and the matching problem. For the problem of vehicle-specific model recognition, a fine model recognition algorithm with improved YOLOv3 algorithm is considered as the detection model, while the RPN module in Mask R-CNN that is further optimized and used for recognition is proposed, and the established local feature dataset is introduced. The hardware system for vehicle model recognition built in this paper is introduced, mainly including database module, file module, feature extraction and comparison module, message queue module, WEB module, and vehicle detection module, and the algorithm proposed in this paper is implanted into the system to verify the practical value of the method. Based on the algorithm development of R-CNN and Faster R-CNN and the design of convolutional layer, the superiority of convolutional neural network in target detection and recognition is illustrated, and the advantages and disadvantages of different methods and network frameworks in target detection are analyzed
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
