Field vehicle control system based on image processing

• Main Authors: LI XINLI, YAO QIUFENG, JIA HONGBIN, LIU SHUAI, ZHENG WEI
• Format: Patent
• Language: Chinese; English
• Published: 30.04.2024
• Subjects: CALCULATING; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

According to the field vehicle control system based on image processing, a special neural network learning model is established to extract a global parameter model suitable for different observation angles and camera parameters, and the global parameter model is used for classifying obstacles and improving the accuracy of obstacle image feature recognition and matching. And calculating external parameters of the vehicle camera according to matching of features of different cameras so as to judge the relationship between the vehicle camera and the obstacle. Therefore, the implementation cost and implementation difficulty of traditional intelligent facilities can be reduced, the real-time performance and safety of control are improved, and intelligent control over the wharf site transport vehicles is achieved. 本发明提出一种基于图像处理的场地车辆控制系统，建立了专门的神经网络学习模型提取适用于不同观察角度、摄像机参数的全局性参数模型，用于对障碍物进行分类，提高障碍物图像特征识别、匹配的准确性。根据不同摄像机特征的匹配，计算车辆摄像机的外部参数，从而判断其与障碍物之间的关系。由此可以降低传统智能设施的实施成本和实施难度，提高控制的实时性和安全性，实现码头场地运输车辆的智能控制。