色彩和人眼视觉系统对图像质量评价的影响研究计算机应用技术专业论文.docxVIP

研究生优秀毕业论文 Abstract Abstract 硕士学位论文 Abstract Images are the main way of humaJl to ullderstaJld the world in their、vork a11d liVes．’Ihe measurement of image fidelit)，and intelligibility is me imponant standard of me process of h啪a11 co鲥tiVe o巧ectiVe things．HoweVer，images may su虢r di虢rent types and degrees of distortion in the image life cycle，leading to me mistakes in the process of image analysis aJld understanding．Obj ectiVe image quality metrics aim to simulate human visual system to realize the measure of digital image quali锣by computers．Therefore，it becomes to be aIl important topic． This paper mainly focuses on the basic problems of obj ectiVe image quali够aSsessment． W．e explore the characteristics of t11e saliency and color space and aJlalyze me import on saliency applied to image quality assessment，and study proper saliency adding strategies． Meanwrhile，we study me innuence of color infomation to image quali够By a11alyzing each color channel aJld making quali够model，we propose a no-reference image qualit)r aSsessment baSed on color info册ation，which proVides pmctical applications of image qualit)，aSsessment． The main contents are summ2Irized as f．ollows： (1)According to impons on saliency印plied to image quali够assessment，a proper saliency adding strategies is proposed．BaSed on the characteristics of the attention mechanism of the human visual system，the innuence of distortion type and degree is exploded．Experimental results demonstrate that me Variation in saliency highly depends on me distortion type and de铲ee．What is more，we propose different saliency adding strategies depended on distortion哆pe．Mean、^，hile，we explore e丘．ects of saliency extmction algoritllms for image quality aSsessment，aIld put fonⅣard some suggestions to choose the proper one． (2)No-reference image q