基于奇异值分解的粮虫图像增强

Enhancing pest image by singular value decomposition

在经典图像增强算法和矩阵奇异值分解的基本理论的基础上,提出了基于奇异值分解的粮虫图像增强算法,该方法没有繁琐的数学变换,实际应用时可以根据需要通过调整高斯噪声的方差来获得不同的图像增强效果,通过增强奇异值矩阵达到增强图像的目的,具有与经典算法相同的功效.通过实验与经典算法对比,Matlab平台的仿真实验表明,该方法对于粮虫图像增强是可行有效的,且简单易行.

The basis of classical image enhancement and matrix singular value decomposition are introduced. Then a novel pest image enhancement method based on singular value decomposition is proposed. Without intricate mathematical transform,the variance of Gaussian noise can be modified to obtain different enhancement results to meet the demands. The proposed method is efficient as the classical algorithm. The experiments on Matlab platform manifest that the proposed approach is efficient,feasible and easy to realize. It′s a practicable enhancement algorithm.