期刊名称：International Journal of Electronics, Communication and Soft Computing Science and Engineering
印刷版ISSN：2277-9477
出版年度：2015
卷号：4
期号：Special 3
出版社：IJECSCSE
摘要：Image Quality assessment (IQA) is used for computationa l models to measure the image quality. In this paper, a new hybrid method of FR (full reference), RR (reduce reference) and NR (no reference) image quality metrics (IQM) is proposed for image quality assessment (IQA ). This proposed method improves the draw backs of FR and RR approaches and works for FR, RR and NR simultaneously. It gives errorless and efficient output. The sparse feature fidelity (SFF) method is used for the FR and RR algorithm as they are similar to each other. The SFF relates the quality o f an image with the reliability to the reference image. The sparse features are acquired by feature detector. It trained on the samples of natural images by an ICA algorithm. The hybrid no reference (HNR) method is used for NR algorithm. HNR model is based on hybrid of curvelet, wavelet and cosine transforms. NR algorithm is used when reference or undistorted images are not available. Natural scene statistic (NSS) properties are used to make NR possible. The peak coordinates of the transformed coefficient occupy well define clusters in peak coordinate space. This method is applicable to arbitrary images without compromising the prediction accuracy of full reference methods. Experimental study shows the proposed method performs better than other previous me thods and improves the effectiveness of IQA
关键词：Image quality assessment (IQA); image qualitymetrics (IQM); full reference (FR); reduce reference (RR); noreference (NR); natural scene statistics (NSS); hybrid no reference(HNR).
