▎ 摘　　要

NOVELTY - The method involves controlling (S1) a first camera module to pre-set first threshold radius for sample image collection of standard battery graphene to obtain a first sample image set. De-noising processing and distance detection are performed (S2) to a first battery image. A convolution kernel is constructed (S3). The convolution is used to check denoising to perform convolution operation. Primary training is performed (S4) to a battery image characteristic set through deep learning. A second camera module is controlled (S5) to pre-set second threshold radius for the sample image collection of the standard battery graphene to obtain a second sample image set. The denoising processing and the distance detection are performed (S6) to a second battery image. USE - Method for judging defect of a battery based on graphene vision. ADVANTAGE - The method enables training image sample obtained by obtaining different distance radius to provide an identification model with accurate identification rate. The method enables providing a pad for subsequent sample collection by comparing pixel points in the camera modules and battery sample material and calculating corresponding distance value. DETAILED DESCRIPTION - An INDEPENDENT CLAIM is included for a system for judging defect of a battery based on graphene vision. DESCRIPTION OF DRAWING(S) - The drawing shows a flowchart illustrating a method for judging defect of battery based on graphene vision. (Drawing includes non-English language text). S1Step for controlling first camera module to pre-set first threshold radius for sample image collection of standard battery graphene to obtain first sample image set S2Step for performing denoising processing and distance detection to first battery image S3Step for constructing convolution kernel S4Step for performing primary training to battery image characteristic set through deep learning S5Step for controlling second camera module to pre-set second threshold radius for sample image collection of standard battery graphene to obtain second sample image set S6Step for performing denoising processing and distance detection to second battery image