Robot visual orientation is an important research topic of mobile robot. A method based on video denoising algorithm for mobile robot visual positioning is proposed in this paper. The motion compensated time-domain video denoising algorithm is used to denoise video images and extract the pixels of the denoising video image. With the camera calibration method, the target space coordinates is determined. The Caiman filter Matlab simulation algorithm is extended to determine the targetlocation. It solved the phenomena of video image smearing and ghosting effectively, and realized the dynamic positioning of targets. The experimental results show that the video denoising effect is good, the camera calibration target is accurate and the target orientation is precise. It achieve the desired effect.%提出基于视频去噪算法的移动机器人的视觉定位方法,应用运动补偿时城视频去噪算法对视频图像去噪,提取去噪视频图像的图像像素,用摄像机标定方法确定目标空间坐标,扩展卡尔曼滤波Matlab仿真算法对目标进行定位,有效地解决了视频图像的拖尾、重影等现象,实现目标动态定位.实验结果表明视频去噪效果良好,摄像机标定准确,目标定位精确度高,达到了预期效果.
展开▼
