基于颜色变化曲线的多视点视频颜色校正【毕业论文】.doc

- PAGE I - 本科毕业设计 （20 届） 基于颜色变化曲线的多视点视频颜色校正 所在学院 专业班级 电子信息科学与技术 学生姓名 学号 指导教师 职称 完成日期 年 月 PAGE 2 PAGE 1 摘　要 【摘要】本文针对多视点视频序列视点间存在的颜色不一致问题，提出了实现基于颜色变化曲线的视点视频颜色校正算法。深入分析视点图像间存在的颜色差异，通过固定比率控制点的方法标注关键帧同时利用Reinhard提出的校正方法实现关键帧的校正，借助于关键帧的统计信息建立以线性内插、抛物线内插和三次样条函数为基础的颜色变化曲线，实现最终的校正效果。基于实验所得的结果，利用主客观评价方法对校正的结果进行评价，实验结果表明，本文实现的颜色校正算法计算复杂度较低，并且在保持图像视觉效果的同时，能较好的消除参考图像和源图像之间的颜色差异，并且能够提高视频的编码性能，证明是一种有效的视频图像颜色校正算法。同时介绍了颜色校正技术的必威体育精装版研究进展状况和本论文存在的不足之处，对颜色校正的研究发展具有一定的意义。 【关键字】多视点视频编码;颜色校正;去相关颜色空间;线性内插;抛物线内插;三次样条函数内插;峰值信噪比（PSNR）。 Abstract 【Abstract】In order to solve non-consistence color appearance between different viewpoint images, We put forward a view video color correction algorithm based on the curve of color. View the color-depth analysis of the existing differences between the images, making key frame by a fixed percentage control points and then achieving the key frames correction using the method proposed by Reinhard. In order to reduce the color error, we can combine with the key frame statistics to establish the color curves between the origin image and the reference image based on the linear interpolation, the parabolic interpolation and the cubic spline interpolation. Based on the experimental results, we can use the subjective and the objective evaluation method to evaluate the calibration results. We show some experiment results to demonstrate the efficiency of our method. Experimental results show that the color correction algorithm to achieve low computational complexity. The schema proposed in our method has exploit the redundancy between color components in different sequences of multi-View. At the same time, the correction can improve the video coding . It proved to be an affection video image color correction algorithm. In this thesis, we also introduce the lastest research progress of the technology and the inadequacy of existing papers. These studies