Processing method and device for distinguishing multiple vehicles from pictures in vehicle insurance case, equipment

Abstract

The invention discloses a processing method and device, for distinguishing multiple vehicles from pictures in a vehicle insurance case and equipment, and relates to the technical field of vehicle picture distinguishing. The processing method for distinguishing multiple vehicles from the pictures in the vehicle insurance case comprises the steps of obtaining first vehicle image information of eachto-be-processed picture; clustering the similarity representation vector corresponding to each vehicle image area to obtain second vehicle image information of each vehicle image area and determininga unique license plate number associated with each clustering group according to all license plate numbers associated with the license plate-containing vehicle image area; and acquiring a license-plate-free vehicle image area in each clustering group, and correspondingly matching each license-plate-free vehicle image area with the unique license plate number associated with the clustering group. Based on a deep learning algorithm, features of vehicle images are extracted, and the similarity between vehicles is measured by using the features. Accurate binding of the vehicle image and the license plate number is realized through a clustering method, and the vehicle distinguishing and recognition efficiency is further improved.

Description

technical field [0001] The invention relates to the technical field of vehicle picture distinction, in particular to a processing method, device and equipment for distinguishing multiple vehicles from pictures in auto insurance cases. Background technique [0002] In the insurance claims industry, there are a large number of vehicle photos that need to be classified and identified in the links of loss determination, loss verification, and claim verification in auto insurance claims. The artificial recognition method is not only slow in recognition speed, but also has the problem of recognition errors. [0003] Aiming at this problem, a solution is proposed in the prior art, which distinguishes different vehicle information through the identification of vehicle color. However, the inventor finds that there are the following problems in the prior art: one is that vehicles with the same or similar colors cannot be accurately Second, the technical solution is to extract color f...

AI Technical Summary

Problems solved by technology

The artificial recognition method is not only slow in recognition speed, but also has the problem of recognition errors

[0003] Aiming at this problem, a solution is proposed in the prior art, which distinguishes different vehicle information through the identification of vehicle color. However, the inventor finds that there are the following problems in the prior art: one is that vehicles with the same or similar colors cannot be accurately Second, the technical solution is to extract color features from the entire photo. When there are multiple cars or a large background in a picture, it is impossible to accurately distinguish the vehicle information; third, the color of some exterior parts of the car, such as the China grid, The lower grille, tires, glass, etc., the color distinction of each vehicle is not obvious. If it is not ruled out, it will affect the recognition accuracy. Fourth, it is impossible to distinguish the license plate number corresponding to each vehicle. Business processes cannot be automated

Images