基于数学形态学的运动图像模糊长度识别方法  

Application of Mathematacal Morphology in Motion Blurred-length Indentification

作　　者：谢晴[1] 玉振明[2] 

机构地区：[1]梧州学院计算机科学系,广西梧州543002 [2]梧州学院信号与信息处理实验室,广西梧州543002

出　　处：《广西科学》2013年第2期125-127,共3页Guangxi Sciences

摘　　要：基于数学形态学思想提出一种运动图像模糊长度识别方法,并进行仿真实验。该方法通过对模糊图像频谱图进行一系列数学形态学的处理,准确提取频谱图的边缘,再利用频谱图中相邻黑色条纹间距与模糊长度的数量关系计算出模糊长度。该方法在35像素范围内能准确识别模糊长度,超过40像素范围时需人工辅助才能识别模糊长度。A method of motion blurred-length identification is proposed based on mathematical morphology. It detects the edge of the spectrum of graph and identifies the blurred-length by a series of mathematical morphology processing. The simulation results show that the proposed method can identify the blurred-length accurately when its length within 35 pixels, whereas it is necessary to artificial assistance in order to identify the blurred-length when the blurred- length is more than 40 pixels.

关 键 词：数学形态学 模糊长度 频谱图 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]