基于统计方法的自适应阈值SUSAN边缘检测方法

Adaptive threshold SUSAN edge-detecting algorithm based on statistical methods

针对传统SUSAN边缘检测方法要人为根据图像不同对比度反复调整设定阈值,检测结果随机性强、不稳定的缺点,提出了一种自适应生成灰度差阈值的改进SUSAN算法.首先,通过统计的方法来反映相邻像素点灰度的空间分布情况;然后,计算图像的对比度,建立对比度与灰度差阈值的关系;最后,生成自适应的灰度阈值,进行边缘检测.本文算法的实验结果与其他边缘检测算法相比,边缘检测效果更好,并且具有抗噪性能.

Aim to the shortcoming of traditional SUSAN algorithm need to set the appropriate gray difference threshold manually,according to the different contrast of the image in specific circumstances,and the edge detection effect is not stable with strong randomness,an improved SUSAN algorithm is put forward with the adaptive gray difference threshold.Firstly,the spatial distribution of adjacent pixels is expressed with statistical method;secondly,the contrast of the image is calculated,the relationship between the contrast and the gray threshold is established;finally,the adaptive gray difference threshold is generated to detect edge.The experimental results show that the performance of the improved SUSAN algorithm is better than other edge-detecting algorithm,and it has more strong anti-noise property.