基于小波分析的足球机器人视觉子系统噪声处理

Noise Disposal for Soccer-Robot Vision Subsystem Based on Wavelet Analysis

在足球机器人视觉子系统中,噪声问题一直是影响比赛成绩的重要因素。由于光照情况对比赛的影响很大,光线的变化在比赛中对机器人的识别会产生干扰,从而影响比赛的胜负。本文对摄像头采集的赛场图像进行小波分析,采用二维离散小波变换处理图像,并选取合适的阈值处理小波系数,最终还原出更加真实的画面,最大程度地减少光线对比赛的影响。实践证明此方法在机器人足球比赛中取得了良好的效果。

In the field of soccer-robot vision subsystem, noise is always a big problem. Since the change of brightness will disturb the identification of robot in the game, the light status always has great influence on the competition. As a result it can affect the achievement of the game. The image of the game field photo collected by a camera was analyzed based on wavelet analysis technique in this paper. By the technique of Two-Dimensional Discrete Wavelet Transform (DWT) and the proper thresholds, more actual image is revived at last and the influence of the light condition on the game is reduced.