Image-based vehicle loss assessment method, device and equipment and storage medium

Abstract

The invention discloses an image-based vehicle loss assessment method, belonging to the technical field of image recognition. The method comprises the following steps of obtaining a color image and acorresponding depth image of the damaged vehicle; using the vehicle part detection model after part training to identify the vehicle part of the color image and confirming the position of each part ofthe vehicle in the color image; splicing the color information of the positions of the identified vehicle parts in the color image and the depth information corresponding to the positions in the depth image into a four-channel data matrix; identifying the four-channel data matrix by using the vehicle damage detection model after the damage identification training, and outputting the identified damage results. The method of the invention combines two kinds of information of color and depth, and greatly improves the classification accuracy of the vehicle damage degrees.

Description

technical field [0001] The invention relates to the technical field of image recognition, and relates to an image-based vehicle damage assessment method, device, equipment and storage medium. Background technique [0002] In the process of auto insurance claim settlement, salesmen are required to assess the damage of the vehicle according to the degree of damage to the vehicle. However, each salesperson has different judgment standards and different levels of experience, which can easily affect the accuracy of vehicle damage assessment. [0003] At present, there are some processing methods in the industry that use the automatic analysis of traffic accident scene images to obtain the classification of preset car damage parts, but they only simply determine the classification of car damage parts, such as the front, side, and rear of the car, and there is no specific method for the equipment. The type of damage, let alone give advice on repairs. The reminder information of th...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to overcome the problem in the prior art that the vehicle damage assessment mainly relies on manual judgment, and proposes an image-based vehicle damage assessment method, device, equipment and storage medium, by combining color images and depth images The above information greatly improves the classification accuracy of the vehicle damage classification model

Images