利用图像自适应轮廓描述模型的插值算法及其Web测试平台-计算机系统结构专业毕业论文.docxVIP

研究生优秀毕业论文 Abstract Abstract 硕士论文 Abstract Image inte叩01alion is a缸rld锄ental question of油age scaling，demosaickiIlg and super res01ution technology，it is also a key teclmology of image editing．How to aVoid zipper e脏ct， reducing the structure and 1uminaIlce e11ror caused by inte叩olation，is the key issues ofthe image interpolation with remaining edge aIld s仃ucture．Funhermore，there is a11 importallt 印plication prospect about CCD imaging deVice for tlle deVelopment a11d lightweight c锄era by emcient image inte叩olation and improVing color filter anr￡Ly(CFA)demosaicking quality． This dissertation starts、vork on ad印tiVe image edge directed ime印olation and CFA demosaicking application，the main achievemems are as follows： (1)0n the basis of researching on the claSsicalimage inte叩olation algorit№，we giVe aIl image interpolation method with keeping the f-eatures of stmcture tenso r．This memod utilizes local stmcture tensor infomation of images for capturing image 10cal edge direction，a11d achieVes adaptiVe structure in inte叩olation process t11rough directional difrusion．Experiments show that mis metllod can reduce the zipper efrect in image interpolation process compared with classicalime印olation methods． (2)For the CFA demosaicking problem，we analyze tlle directionallimitations of classic interpolation in demosaicking，and research a demosacking method by local directional interp01ation and nonlocal adaptiVe thresholding．The contrastiVe experiment results indicate that edhancing the anisotropy and directional adaptabili妙in image inte叩01ation can improVe the Visual quality of the demosaicking image． (3)We study the mechanism of the local image adaptiVe comour stencil description model， and giVe constmction method of the 10cal adaptiVe contour stencil by using