Vehicle type identification method and system based on deep neural network

Abstract

In order to improve the efficiency of vehicle positioning to meet the real-time requirements for vehicle type identification, the present invention provides a vehicle type identification method and system based on a deep neural network. The method comprises: integrating candidate target extraction and target identification into a network, and using end-to-end detection / identification methods to integrate feature extraction, target location, and target detection into a single network. According to the method and system provided by the present invention, target extraction t is no longer extracted from the original image, but is extracted from the high-dimensional feature map with very small dimensions by using the reference point and multi-dimensional coverage manner, and at the same time, the method for sharing the deep convolutional network parameters is used in the feature extraction layer, so that repetitive calculation of features is avoided, the identification efficiency is greatlyimproved, 20fps is reached in reality, the effect of the processing efficiency of a single server reaches 2 million sheets / day, and the requirement for real-time vehicle type identification is satisfied.

Description

technical field [0001] The invention belongs to the field of intelligent transportation, and relates to a vehicle identification method and system based on a deep neural network. Background technique [0002] For a long time, video analysis technology, especially license plate recognition technology, has effectively promoted the rapid development of various fields such as smart transportation and safe cities, and provided strong technical support for city managers to quickly retrieve vehicles, track tracking, early warning interception and comprehensive judgment . However, license plate recognition technology has the following disadvantages: [0003] 1. The efficiency and completeness of vehicle retrieval cannot meet the requirements. Especially in case detection, due to the interference of objective factors such as too small targets, changing light, and scene constraints, as well as human factors such as false plates, fake plates, and defaced / occluded / missing license plat...

AI Technical Summary

Problems solved by technology

[0003] 1. The efficiency and completeness of vehicle retrieval cannot meet the requirements

Especially in case detection, due to the interference of objective factors such as too small targets, changing light, and scene constraints, as well as human factors such as false plates, fake plates, and defaced / occluded / missing license plates, retrieval based on license plate information The method often cannot quickly lock the target vehicle at the first time, causing the investigators to lose the best opportunity to solve the case, which brings great difficulty to the case detection

[0004] 2. Lack of an effective solution for real-time identification of vehicle models

[0007] 1. Background modeling and moving target detection: a large amount of calculation, poor illumination robustness, and it is difficult to establish an effective background at night due to too dark light

[0008] 2. Sliding window traversal method: Pyramid scaling of the original image is required, and there are a large number of repeated calculations of image region features between adjacent sliding windows, which is very time-consuming and difficult to achieve real-time performance

[0009] 3. Image segmentation: use the image segmentation algorithm to segment the target area and then send the segmented area to the classifier for recognition. After segmentation, the number of targets is too large and there are many repeated areas, which still cannot satisfy real-time performance.

Images