Design of Tracking Smart Car Based on LabVIEW

Abstract

This paper proposes a trajectory recognition and tracking scheme based on machine vision, and on this basis, designs a tracking smart car suitable for most scenes with trajectory belts. A deflection calculation method based on the geometric position of the pixel is proposed. Firstly, the original image of the trajectory belt is segmented by adaptive threshold using the between-class variance method to obtain a binary image, and then the trajectory deflection angle is calculated by obtaining the geometric center point coordinates of the partial image of the trajectory belt. Compared with some existing tracking algorithms, this method can better adapt to different environments, and guarantee the real-time and accuracy of processing while occupying less computing resources. Experimental results show that this method can complete path recognition and tracking tasks, and it has a better tracking effect when the trajectory is curved.