基于融合计算机视觉图像的差异特征识别  

Difference Feature Recognition of Computer Vision Image Based on Fusion

作　　者：李菊霞[1] 

机构地区：[1]山西农业大学信息科学与工程学院,山西太谷030801

出　　处：《计算机仿真》2015年第1期442-445,共4页Computer Simulation

摘　　要：融合图像的特征识别是模式识别领域重要的研究方向。在图像差异特征识别过程中,一旦图像质量较差,造成无法根据大量图像特征趋于同质化,传统的识别方法是针对同质化特征很难建立准确的识别模型,效果不佳。提出多小波变换融合的图像差异特征识别方法。通过分解小波N层,取得图像差异特征的频率域,选择相应的融合规则取得分解合成图像的多分辨率,重构的融合图像,根据分离的差异特征完成图像识别。实验结果表明,利用改进算法进行图像差异特征识别,能够极大地提高识别的准确性,满足图像处理的实际需求。The characteristics recognition of the fusion images is an important research direction in the field of pattern recognition. In the process of image difference feature recognition, once image quality is not good, a lot of image features will tend to be more homogeneous. Traditional recognition methods are difficult to establish accurate recognition model based on the homogeneity, and the recognition results are not satisfied. A difference feature recognition method of fusion images is presented based on multiple wavelet transform. By decomposing wavelet n-tier, the image differences characteristics in the frequency domain is decomposed, the fusion rules in the difference of frequency domain are selected to obtain the synthesis image multi-resolution decomposition, the fusion image is rebuilt, and the image recognition is completed based on the differences of separation characteristics. The experimental results show that the improved algorithm can greatly improve the accuracy of recognition and meet the practical requirements of image processing.

关 键 词：视觉图像 融合 识别 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]