US 11,790,626 B2
Method and apparatus for recognizing bitting code of key, and storage medium and image collection device
Yongfeng Xi, Guangdong (CN); Yuan He, Guangdong (CN); Yijie Hao, Guangdong (CN); Guozhong Cao, Guangdong (CN); Chenglong Li, Guangdong (CN); Guoming Huang, Guangdong (CN); and Shuli Xi, Guangdong (CN)
Assigned to SHENZHEN XHORSE ELECTRONICS CO., LTD, Guangdong (CN)
Appl. No. 17/439,363
Filed by SHENZHEN XHORSE ELECTRONICS CO., LTD, Guangdong (CN)
PCT Filed Aug. 27, 2020, PCT No. PCT/CN2020/111755
§ 371(c)(1), (2) Date Sep. 14, 2021,
PCT Pub. No. WO2021/143146, PCT Pub. Date Jul. 22, 2021.
Claims priority of application No. 202010056497.4 (CN), filed on Jan. 18, 2020; and application No. 202020110251.6 (CN), filed on Jan. 18, 2020.
Prior Publication US 2022/0189135 A1, Jun. 16, 2022
Int. Cl. G06T 7/73 (2017.01); G06V 10/141 (2022.01); B23C 3/35 (2006.01); G06V 10/44 (2022.01); G06V 10/145 (2022.01)
CPC G06V 10/141 (2022.01) [B23C 3/355 (2013.01); G06T 7/73 (2017.01); G06V 10/145 (2022.01); G06V 10/44 (2022.01); B23C 2235/41 (2013.01); G06T 2207/30136 (2013.01)] 19 Claims
OG exemplary drawing
 
1. A key tooth profile code recognition method, wherein the key tooth profile code recognition method comprises:
using a back light source and/or side light sources to acquire an image of a key to be recognized; and
recognizing a tooth profile code of the key to be recognized, based on the image, wherein the recognizing a tooth profile code of the key to be recognized, based on the image comprises:
performing an image correction on the image to obtain a corrected image after correction, wherein the image correction comprises:
association relationship between a key type and the perspective feature point information corresponding to the key type is pre-established, perspective feature points in the acquired image are firstly recognized based on the key type, and the original coordinates of the perspective feature points in the original image are obtained, the perspective transformation matrix is calculated according to the original coordinates of the perspective feature points and the set target coordinates, and then the perspective transformation matrix is used to perform the perspective transformation processing on the acquired image, so as to obtain a corrected image after the perspective transformation processing, then the tooth profile features are extracted from the corrected image to obtain the tooth profile code of the key to be recognized;
when there are multiple images acquired by the camera light sources in different orientations as for the same key to be recognized, the perspective transformation matrix is calculated based on one of the images, and then the same perspective transformation matrix is used to perform the perspective transformation processing on multiple images of the same key to be recognized, so as to obtain multiple corrected images on the same plane and with the same transformation specifications; and
determining the tooth profile code of the key to be recognized based on the corrected image.