【摘要】 提出了一种小波边缘提取算法,应用多孔算法对视频信息进行小波分析,基于模极大值类内MIN variance原理,推导出高低2个阈值,进行候选边缘点的筛选。对在2个阈值之间的模极大值进行比较,进一步筛选候选边缘点。该方法不需要人为的设定任何系数或者初始值,根据不同的视频信息自适应地确定双阈值,避免了反复调试阈值系数,实验结果表明了该方法的有效性。更多还原

【Abstract】 In this paper,a wavelet algorithm based on adaptive double threshold for edge detection is developed and a trous algorithm is used to transform the visual information of an image and two thresholds are derived for selecting the edge points based on modulus maximum and minimum interclass variance.Then the edge points are further screened by comparing the modulus maximum between those two thresholds.The proposed algorithm can adaptively determine the double thresholds according to different images and avoid artificial coefficients setting.The subsequent experiments results prove the effectiveness of the proposed algorithm.更多还原