A real-time radar detection area detection method

Abstract

The invention provides a real-time radar detection area detection method, which belongs to the field of image processing. The present invention uses a deep neural network combined with a traditional image processing method to perform semantic segmentation on the robot's visual image. In the first stage, the segmentation based on the deep neural network is performed to determine the semantic level of each pixel on the image. In the second stage, the OTSU threshold algorithm is used to perform threshold segmentation on the image. The invention judges the radar detection area while distinguishing the walkable area and the non-walkable area, realizes the real-time detection of the radar detection area in machine vision, and facilitates the combination of computer vision technology's image perception information and laser radar detection. It has played a practical role in machine vision image segmentation, machine vision road recognition, and machine vision obstacle detection.

Description

technical field [0001] The invention belongs to the field of image processing, in particular to a real-time radar detection area detection method. Background technique [0002] In recent years, with the rapid development of computer software and hardware technology, electronic information technology, and control technology, robot technology has also made great progress, from the workpiece processing robot that initially worked on the factory assembly line to the service robot that interacts with customers. Security robots for security inspections, disaster rescue robots for extinguishing fires, robot technology is used in various fields such as service industry, processing industry, military, etc., and presents a development trend of integration, platformization, and intelligence. Among all kinds of robots, the security inspection robot is an indispensable part, and the technology applied to the security inspection robot is also very complicated, involving control theory, ma...

AI Technical Summary

Problems solved by technology

However, since image perception is not absolutely accurate in judging the robot's front vision, when computer vision technology perceives robot vision images, in addition to dividing the walkable area and the non-walkable area, the feasibility of some areas is to be determined. Further scanning and detection are required to effectively avoid various obstacles on the way to patrol

Images