Embedded monocular vision navigation system based on navigation line recognition

Abstract

An embedded monocular visual navigation system based on navigation line recognition, the system includes the following processing steps: the first step, video signal acquisition, the single-chip microcomputer carries out video signal acquisition; the second step, distortion correction, the image collected by the camera is processed Vertical and horizontal distortion correction; the third step, path recognition and extraction, separating and extracting the navigation line from the background image; the fourth step, steering control, building a steering model to calculate the relationship between the turning radius of the smart car and the steering gear swing angle , to provide data for the steering control of the smart car; the fifth step, steering gear control, to realize the smart car moving along the navigation line; the sixth step, speed control, using the PID control method to control the running speed of the smart car; the present invention collects and navigates with a camera The line information has good visual foresight, which effectively improves the performance of the navigation system. The invention realizes image dynamic threshold calculation and self-adaptive adjustment in complex environment, and has better stability under complex light.

Description

technical field [0001] The patent of the present invention relates to the field of locomotive automatic control, specifically an embedded monocular vision navigation system that can realize the autonomous navigation function of vehicles following ground navigation lines. Background technique [0002] With the development of science and technology and economy, autonomous guided vehicles (AGV) have been widely used in the industrial field. The vision-based navigation system has become a research hotspot in the field of AGV at home and abroad. Although the current PC-based visual navigation system occupies the mainstream position, there are still many fatal flaws; the embedded visual navigation system has become an inevitable trend to replace the PC board system. However, due to the limitation of software / hardware resources, the adaptability of embedded visual navigation system to complex environments is very weak, which seriously restricts the popularization and application o...

AI Technical Summary

Problems solved by technology

[0003] In order to solve the problem that the current embedded visual navigation system is very weak in adaptability to complex environments and limited by software / hardware resources, the present invention uses a CCD camera as a visual sensor to realize image dynamic threshold calculation and adaptive adjustment for complex environments, and based on The method of real-time acquisition and distortion correction of long-distance image information with limited resources realizes an embedded monocular vision navigation system that operates stably in a variable light environment and has the ability to obtain long-distance information

Images