Unlock instant, AI-driven research and patent intelligence for your innovation.

Checkerboard check method based on deep learning semantic segmentation

A semantic segmentation and deep learning technology, applied in the field of automotive driving assistance, can solve the problems of easy false detection, difficult to eliminate, easy to be interfered by other objects, etc., and achieve high precision.

Active Publication Date: 2021-07-13
SAIC MAXUS AUTOMOTIVE CO LTD
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But because it is the corner point of the checkerboard that is directly detected, there is no other verification information. When the corner point given by deep learning is wrong, it is difficult to rule it out.
At the same time, the checkerboard corner feature is easily disturbed by other objects in the physical world. Sometimes the corner of a stain on the ground may also be considered as a checkerboard corner point, so this method is prone to false detection.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Checkerboard check method based on deep learning semantic segmentation
  • Checkerboard check method based on deep learning semantic segmentation
  • Checkerboard check method based on deep learning semantic segmentation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0031] The present invention will be further described below in conjunction with the accompanying drawings.

[0032] The present invention is a checkerboard detection algorithm with strong adaptability to the environment and high robustness. The scheme is designed by combining the relatively strong adaptability of deep learning to the environment and the high robustness of OCamCalib in the checkerboard check.

[0033] The realization of this scheme is mainly divided into the following steps:

[0034] Use the method of deep learning semantic segmentation to segment the checkerboard in the image from the image. Implementing this step includes selecting the segmentation network, GT (groudture) labeling, network parameter training, and final prediction.

[0035] For the segmentation network, you can choose the commonly used DeepLab, etc., or you can modify multiple networks such as resnet to realize the segmentation function. It should be noted that some target checkerboards onl...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to a checkerboard inspection method based on deep learning semantic segmentation; the method is used for checkerboard inspection during calibration of any camera; a checkerboard image is firstly acquired to achieve the purpose of calibrating internal and external parameters of the camera, and the image is acquired through a vehicle-mounted camera and checkerboard inspection is performed. Compared with a traditional image processing method, the invention is high in accuracy and small in error degree, can be combined with the advantages of traditional checkerboard detection, improves the adaptability to interference items in the environment, and also improves the adaptability to environment illumination.

Description

technical field [0001] The invention relates to a checkerboard inspection method based on deep learning semantic segmentation, which belongs to the technical field of automobile driving assistance. Background technique [0002] At present, the camera-based vision system plays an increasingly important role in intelligent driving vehicles. Cameras are similar to the eyes of a smart driving car and are used for object detection to judge the surrounding environment. After the target is detected, the distance from the target to the car needs to be known, which can be obtained by calculating the pose of the camera relative to the car. The pose of the camera relative to the car needs to be calibrated. [0003] Camera calibration mainly refers to the calculation of the internal and external parameters of the camera (relative to the pose of the car). At present, the most commonly used method of camera calibration is to calculate the internal and external parameters of the camera ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06T7/80G06N3/04G06N3/08
CPCG06T7/80G06N3/08G06T2207/20081G06T2207/20084G06N3/045
Inventor 杨盼王瑞罗先伟朱晶星赖杰
Owner SAIC MAXUS AUTOMOTIVE CO LTD