Research on Obstacle Avoidance of Intelligent Driving Vehicles Based on Monocular Camera

Obstacle-avoidance control is a key issue of self-driving vehicles, and the object-detection algorithm is the basis of obstacle-avoidance control. Therefore, this paper proposes an object detection algorithm using a monocular camera as the sensor for obstacle-avoidance system in intelligent driving vehicles. The basic principles of object detection algorithm based on monocular camera are introduced in this paper, based on which, research on obstacle avoidance is done. Histogram of oriented gradient (HOG) features and support vector machine (SVM) are used to detect and classify the obstacles in front of the ego vehicle using a camera. Sliding windows are utilized to detect the targets. Then, we calculate a heatmap and label them to remove duplicate detection frames. Simulation and experiments are implemented based on the data processed by the camera and results show the robustness of the proposed algorithm, and its effectiveness in detecting and avoiding the obstacles.