基于改进型Canny算子的虹膜图像外边缘检测

Iris Image Edge Detection Based on Improved Canny Operator

基于使用传统Canny算子检测虹膜图像后,部分图像抛物线和圆的检测出现细节不完整等问题,提出一种基于自适应动态阈值的算子。将整幅图像分割为若干个子区域,令子图像之间有一定区域的重叠,然后依据非极大值抑制之后的输出结果自适应地选取每个子图像的高、低两个阈值。为了验证改进算法的有效性,因而对CASIAI虹膜数据库进行实验并对结果进行分析,得出改进Canny算法相比传统Canny算子不仅计算复杂度有所降低,而且能够更好地检测边缘信息。

In this paper, an arithmetic which is based on adaptive dynamic threshold is put forward. It is directed against the traditional Canny operator detection used in the iris image, which has been confronted with incomplete problems in part of its image parabola and round inspection details. This arithmetic segments the whole image into several sub areas in order to make a certain area of overlap between the sub images. After that, the output of the high and low two thresholds of each sub image will be selected adaptively based on the maximum inhibition. This paper is to verify the validity of the improved arithmetic so that it tests the CASIAI iris database and analyzes the results. It makes a conclusion that compared with the traditional Canny operator, the improved Canny algorithm not only has reduced computing complexity, and it is better capable to detect the edge information.