CAJ | 학술논문

为了有效减少噪声对运动目标检测的影响，提出了一种利用均值漂移聚类实现运动目标检测的方法。首先运用MeanShift算法分别对三帧连续图像进行平滑去噪处理，然后对图像进行边缘提取，最后通过三帧差分法对三帧图像进行差分，进而得到运动目标。实验结果表明，该方法可以有效地抑制噪声并提取出运动目标。
위료유효감소조성대운동목표검측적영향，제출료일충이용균치표이취류실현운동목표검측적방법。수선운용MeanShift산법분별대삼정련속도상진행평활거조처리，연후대도상진행변연제취，최후통과삼정차분법대삼정도상진행차분，진이득도운동목표。실험결과표명，해방법가이유효지억제조성병제취출운동목표。
In order to reduce the impact of noise on moving target detection effectively. This paper presents a Mean Shift clustering method to detect the moving target. Firstly, noise in three consecutive frames image is removed by using the Mean Shift algorithm, tThen, image edge extraction is done, and the three-frame-differencing method is used for the three consecutive frames image to get the moving object. Experimental resuhs show that this method can restrain the noise and extract the moving target effectively.