一种非线性降维改进算法在图像处理中的应用

摘要

The structure of high dimensional images with high recognition rate of Viti Levu down the key link, and some of the traditional algorithm to reduce the dimensions of the image processing has yielded some results, but exposed the shortcomings of their own. In order to achieve the high ideals of nonlinear effects of image recognition, dimension reduction in the traditional algorithm analysis and refinement of their advantages, put forward an improved algorithm for nonlinear dimensionality reduction to solve the shortcomings of the traditional algorithm. Respectively, in the ORL and CMU PIE database, the simulation experiment on the image to show that the algorithm for high-dimensional image pattern recognition viability.%对高维非线性结构的图像进行降维是提高识别率的关键环节,而一些传统的算法在对图像进行降维处理过程中虽然取得了一定的成效,但也暴露了它们各自的缺陷.为了达到对高维非线性图像识别的理想效果,在对传统的降维算法分析并对其优势进行提炼的基础上,提出了一种改进的非线性降维算法,解决了传统算法的缺陷,并分别在ORL和CMU PIE数据库图像上进行了仿真实验,从而验证了该算法对高维非线性图像在模式识别上的可行性.