A car model fine classification system based on convolutional neural network

Abstract

The invention discloses a vehicle fine classification system based on a convolutional neural network, which provides a method for classifying a large number of vehicle images whose features are not obvious and difficult to extract, and obtains the brand and model information of the vehicle through image analysis without using a license plate Information that is easy to be blocked and tampered with, such as vehicle logos, can be used in traffic supervision, traffic inspection, traffic statistics, criminal investigation and other fields of licensed vehicles, unlicensed vehicles, and covered vehicles; it solves the problem of using images for vehicle brands and models Super large number of classification problems; use a hybrid convolutional neural network with multiple sub-networks to input the local and global pictures of the vehicle into each sub-network at the same time, such as inputting the face picture, body picture and rear picture of the vehicle into different sub-networks The network is scored through the subnet scoring layer to obtain classification results.

Description

technical field [0001] The present invention provides a fine classification system for vehicle models based on convolutional neural network, which inputs vehicle face pictures, body pictures, and vehicle rear pictures into a hybrid convolutional neural network with multiple sub-networks to complete the classification of vehicle brands and models , targeted design, training, and tuning to achieve relatively complete vehicle information acquisition and high vehicle recognition accuracy, which belongs to the field of image processing technology. Background technique [0002] In recent years, with the development of the economy and the progress of the automobile industry, the number of cars in cities has increased year by year, followed by a series of serious problems such as traffic accidents, traffic congestion, and traffic pollution. In order to create a safe, convenient, economical and efficient transportation environment, developed countries have invested a lot of money and...

AI Technical Summary

Problems solved by technology

Such a classification cannot meet the requirements of traffic supervision, traffic inspection, and traffic statistics in intelligent transportation.

Images