曾恬恬,吴建清*.基于MATLAB的图像拼接技术及其实现[J].海南师范大学学报自科版,2021,34(1):40-48 |
基于MATLAB的图像拼接技术及其实现 |
Image Mosaic Technology and Its Implementation Basedon MATLAB |
|
DOI:10.12051/j.issn.1674-4942.2021.01.007 |
中文关键词: 图像拼接 SURF 加权平均 分辨率 重复率 |
英文关键词: image mosaic SURF weighted average resolution repetition rate |
基金项目:海南省高等学校科学研究项目(Hjkj2013-23) |
|
摘要点击次数: 1165 |
全文下载次数: 1461 |
中文摘要: |
基于特征点SURF算法的图像配准和加权平均融合算法,将有重叠区域的两张图像拼
接成一张高分辨率的全景图像,并运用MATLAB软件实时仿真拼接图像。首先利用SURF算法匹
配图像,经仿射变换后得到含有拼接缝的合并图像;再分别对比不同分辨率、重复率的三组测试数
据;接着使用两种融合算法融合比对最优图像;最后设计了一个简单的GUI界面实现图像拼接。
实验表明:分辨率越高,图像信息量越大,匹配效果就越清晰,但会耗费更多的运算时间;重复率增
加,有效匹配点占比增大,得出图像重复率在40%及以上时会使匹配更精准。加权融合算法能更
好地实现图像之间的平滑过渡。 |
英文摘要: |
Image mosaic was used to splice two images with overlapping areas into a high-resolution panoramic image,
based on the image registration of the feature points SURF algorithm and weighted average fusion algorithm, and real-time
simulation was conducted by using MATLAB software. First, the SURF algorithm was used for image matching, and a
merged image with a stitching seam was obtained after affine transformation. Three sets of data with different resolutions
and repetition rates were selected for test and comparison, then two fusion algorithms were used to fuse and compare the op‐
timal image. Finally, a simple GUI interface was designed to realize image mosaic. Experiments show that the higher the
resolution,the greater the amount of image information and the clearer the matching effect, but more operation time was
needed. With the repetition rate increased, the proportion of effective matching points increased. It is concluded that the
matching was more accurate when the repetition rate of the image was more than 40%. The weighted fusion algorithm can
be more effective to achieve the smooth transition between images. |
查看全文
查看/发表评论 下载PDF阅读器 |
关闭 |