混合像元分解下二维激光图像亚像素边缘细化研究

Sub pixel edge thinning of two dimensional laser image under mixed pixel decomposition

针对传统方法边缘细化不准确、效率低的问题,提出混合像元分解下的二维激光图像亚像素边缘细化方法。首先通过非局部均值算法对二维激光图像进行去噪,提取二维激光图像亚像素边缘特征,然后对二维激光图像非极大值进行抑制,引入自适应加权平均融合法,最后结合反扩散思想法,实现对二维激光图像亚像素边缘的细化。结果表明,本文方法的细化结果相比传统细化方法在细化误差、信噪比方面均有一定的提高。

In view of the defects of the traditional edge thinning,such as the inaccuracy and low efficiency,a new sub pixel edge thinning method of 2D laser images under mixed pixel decomposition is designed. First denoising the 2D laser images by non-local means algorithm to extract its edge features. Then do the non maximum suppression of it and introduce adaptive the weighted average fusion method. Finally realize the sub pixel edge thinning of 2D laser images.The experimental results show that the this proposed method is improved in the thinning error rate and SNR than traditional way.