Glass recognition method using monocular structured light camera and monocular structured light camera
By using stereo registration and fitting function to complete the depth map, combined with polarizer technology, the problem of inaccurate identification of glass obstacles by monocular structured light cameras was solved, ensuring the accuracy of the robot obstacle avoidance system.
CN115457153BActive Publication Date: 2026-07-14SHENZHEN BERXEL PHOTONICS CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHENZHEN BERXEL PHOTONICS CO LTD
- Filing Date
- 2022-10-24
- Publication Date
- 2026-07-14
AI Technical Summary
Technical Problem
When encountering glass obstacles, monocular structured light cameras cannot accurately identify the glass and the objects behind it, leading to misjudgments by the obstacle avoidance system.
Method used
The first and second monocular devices are registered using a stereo registration method. The depth map is completed by fitting a function, and the reflected light from the glass is removed by a polarizer to identify objects behind the glass.
Benefits of technology
It enables accurate glass identification, providing a reliable data foundation for robot obstacle avoidance.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN115457153B_ABST
Abstract
The application discloses a monocular structured light camera recognition method and monocular structured light camera. The method comprises the following steps: a first monocular device and a second monocular device are registered through a stereoscopic registration mode to obtain a registered first monocular device and a registered second monocular device; a depth map of pure glass is obtained according to a depth map collected by the registered first monocular device and the registered second monocular device; and a depth map of an object behind the glass is obtained by complementing the depth map of the pure glass through a fitting function. The application can accurately recognize the glass to provide a data basis for subsequent obstacle avoidance of a robot.
Need to check novelty before this filing date? Find Prior Art