CAJ | 학술논문

在同一场景不同光线下得到的图像序列,常出现过曝光或曝光不足的现象,造成图像质量低,重要信息丢失.为解决这些问题,文中对拉普拉斯金字塔与梯度金字塔变换融合的方法进行了分析研究,后者增加了图像方向信息.在图像融合过程中,采用分块的方法,首先生成原始多曝光图像的权重图,再根据金字塔原理对权重图进行加权平均得到融合后图像.实验结果表明,该算法能够得到具有更多有用信息的高对比度的融合图像,融合效果良好,具有较好的实用价值.
재동일장경불동광선하득도적도상서렬,상출현과폭광혹폭광불족적현상,조성도상질량저,중요신식주실.위해결저사문제,문중대랍보랍사금자탑여제도금자탑변환융합적방법진행료분석연구,후자증가료도상방향신식.재도상융합과정중,채용분괴적방법,수선생성원시다폭광도상적권중도,재근거금자탑원리대권중도진행가권평균득도융합후도상.실험결과표명,해산법능구득도구유경다유용신식적고대비도적융합도상,융합효과량호,구유교호적실용개치.
The images which obtained in the same scene under different light are often over exposed and underexposed. These image have low quality. Lots of important information is loss. To solve these problems,the Laplacian pyramid and gradient pyramid transformation are analyzed. The latter increases the information of image direction. In the process of image fusion,using the block method,it is first to gen-erate the weight diagram of the original multi-exposure images. Then generate a high dynamic range images based on Laplacian pyramid weighted average of the weight figure. The experimental results show that the algorithm can obtain high-contrast fusion image with more useful information. The algorithm works well and has good practical value to some extent.