一种基于CNN的运动对象分割算法

A Motion Object Segmentation Method Using Cellular Neural Network

提出一种基于细胞神经网络的运动对象分割算法。通过引进除噪模板和改进边缘检测模板,去除了图像采集和处理过程中的噪声影响,得到比较满意的图像边缘信息。实验结果验证了算法的有效性,而且处理速度与图像大小无关,能够实现图像实时处理。

Based on cellular neural networks（CNN）,a moving objects segmentation method are proposed.By introducing noise removing templates and improving edge detection templates,which remove the noise of acquisition and processing,and satisfying information about image edge is got.Experimental results demonstrate the effectiveness of the method,such as calculation speed independent of image size,real-time performance.