摘要
针对足球机器人比赛中视觉系统涉及的图像背景分离难点,进行了自适应背景图像分离算法的研究,采用自组织神经网络与连续影像相减的背景分离方法,按所设计的算法进行预处理,仿真了分离过程,并应用在足球机器人视觉系统上,提高了机器人定位的可靠性和准确性。
Aiming at the problem of separating the image background in the game of soccer robot vision, research is carried out by adopting the adaptiwe background image separation algorithm. The method of separating bound such as a self-organizing neural networks and continuous image subtraction was adopted. According to the predesigned algorithm processing, the separation process was simula- ted, and application of the separation algorithm to soccer robot vision system was done. The accuracy and reliability of robot positioning are improved.
出处
《机床与液压》
北大核心
2017年第19期110-114,共5页
Machine Tool & Hydraulics
关键词
图像分离
神经网络
影像相减
Image segmentation
Neural network
Image subtraction
