Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020140371 - DEEP LEARNING-BASED VEHICLE DAMAGE IDENTIFICATION METHOD AND RELATED DEVICE

Publication Number WO/2020/140371
Publication Date 09.07.2020
International Application No. PCT/CN2019/088801
International Filing Date 28.05.2019
IPC
G06K 9/62 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
62Methods or arrangements for recognition using electronic means
CPC
G06K 9/46
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
36Image preprocessing, i.e. processing the image information without deciding about the identity of the image
46Extraction of features or characteristics of the image
G06K 9/62
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
62Methods or arrangements for recognition using electronic means
G06N 3/04
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
04Architectures, e.g. interconnection topology
G06N 3/08
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
08Learning methods
G06Q 40/08
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
40Finance; Insurance; Tax strategies; Processing of corporate or income taxes
08Insurance, e.g. risk analysis or pensions
Applicants
  • 平安科技(深圳)有限公司 PING AN TECHNOLOGY (SHENZHEN) CO., LTD. [CN]/[CN]
Inventors
  • 石磊 SHI, Lei
  • 马进 MA, Jin
  • 王健宗 WANG, Jianzong
Agents
  • 广州三环专利商标代理有限公司 SCIHEAD IP LAW FIRM
Priority Data
201910015378.104.01.2019CN
Publication Language Chinese (ZH)
Filing Language Chinese (ZH)
Designated States
Title
(EN) DEEP LEARNING-BASED VEHICLE DAMAGE IDENTIFICATION METHOD AND RELATED DEVICE
(FR) PROCÉDÉ D'IDENTIFICATION D'ENDOMMAGEMENT DE VÉHICULE BASÉ SUR UN APPRENTISSAGE PROFOND ET DISPOSITIF ASSOCIÉ
(ZH) 基于深度学习的识别车辆损伤的方法和相关装置
Abstract
(EN)
The present application provides a deep learning-based vehicle damage identification method and a related device. The method comprises: obtaining a first image corresponding to a target vehicle, the target vehicle being a vehicle with a damage to be identified, and the first image being an image comprising a damaged part of the target vehicle; processing the first image by means of a residual dense network to obtain a second image, the resolution of the second image being higher than that of the first image; detecting the second image by means of a damage detection model based on a single-point multi-box detector algorithm to obtain first information, the first information comprising position coordinates of the damaged part in the second image; and marking, according to the position coordinates, an area where the damaged part is located in the second image. According to the technical solution, the tiny damage of the vehicle can be identified, and the vehicle damage identification precision is improved.
(FR)
La présente invention se rapporte à un procédé d'identification d'endommagement de véhicule basé sur un apprentissage profond et à un dispositif associé. Le procédé comprend les étapes consistant à : obtenir une première image correspondant à un véhicule cible, le véhicule cible étant un véhicule présentant un dommage à identifier, et la première image étant une image comprenant une partie endommagée du véhicule cible ; traiter la première image au moyen d'un réseau dense résiduel pour obtenir une seconde image, la résolution de la seconde image étant supérieure à celle de la première image ; détecter la seconde image au moyen d'un modèle de détection de dommages sur la base d'un algorithme de détecteur multi-boîtier à point unique pour obtenir des premières informations, les premières informations comprenant des coordonnées de position de la partie endommagée dans la seconde image ; et marquer, en fonction des coordonnées de position, une zone où se trouve la partie endommagée dans la seconde image. Selon la solution technique, les petits dommages du véhicule peuvent être identifiés, et la précision d'identification des dommages du véhicule est améliorée.
(ZH)
本申请提供基于深度学习的识别车辆损伤的方法和相关装置,其中,所述方法包括:获取目标车辆对应的第一图片,所述目标车辆为待识别损伤的车辆,所述第一图片为包含所述目标车辆的损伤部位的图片;通过残差密集网络对所述第一图片进行处理,得到第二图片,所述第二图片的分辨率高于所述第一图片的分辨率;通过基于单点多盒检测器算法的损伤检测模型对所述第二图片进行检测,得到第一信息,所述第一信息包括所述损伤部位在所述第二图片中的位置坐标;根据所述位置坐标在所述第二图片中标记出所述损伤部位所在的区域。该技术方案可以识别车辆的微小损伤,提高车辆损伤识别的精度。
Also published as
Latest bibliographic data on file with the International Bureau