The path recognition algorithm in autonomous driving technology is crucial in controlling the navigation and tracking of vehicles. In this article, a simulated road is constructed in the laboratory using a model car. Firstly, images are captured using a camera and a TC264DA chip is used to process and recognize the path of the acquired images. The TC264DA has a high main frequency of up to 200MHz and powerful performance, making it capable of completing data processing and vehicle control processes. Secondly, the images are preprocessed and the optimized OTSU algorithm is used to calculate the maximum inter-class variance of the grayscale in the entire image, and the image is binarized. Then, pixel filtering is applied to remove black noise in the lane. Finally, after the filtering process is completed, the lane boundaries are extracted and fitted to obtain the mid-line of the road, ensuring that the vehicle travels along the mid-line and is able to automatically recognize relevant paths, enabling the vehicle to select appropriate routes in complex road conditions.