图像拼接技术在茶树病虫害防治研究中的应用
Application of Image Mosaic Technology in the Study of Tea Plant Diseases and Insect Pests Control
罗辉兰 1朱兰艳1
作者信息
- 1. 昆明理工大学国土资源工程学院,云南昆明650000
- 折叠
摘要
为较好地实现茶树全景图像的拼接,针对图像拼接过程中存在同名点检测时间长、非同名点基数大的问题,提出一种顾及排列顺序的图像拼接算法.该算法以基准图像为起始,采用相关性原理,以矩形块为计算单位,提取图像间相关区域;后序图像区域提取基于前序图像区域位置计算;最后在目标区域中完成同名点检测,并实现特征匹配、图像配准与融合.试验结果表明,顾及排列顺序的图像拼接算法基于目标区域提取同名点能有效地减少非同名点数量,缩短同名点查找以及特征匹配时间,能够有效提高图像配准速度并实现茶树全景图像的无缝拼接.
Abstract
To achieve a better panoramic image stitching of tea plant, in view of the fact that the process of image mosaic exists the problem of having long time of corresponging image points detection and little corresponging image points, an image mosaic algorithm considering the sequence was proposed.This algorithm was starting with the reference image and correlation theory was used to extract correlation areas with rectangular-block as unit;the extracting of post sequence image correlation area was based on the location of pre sequence image correlation area;finally, corresponging image point in target region was detected, and then features matching, image registration and fusion were implemented.The experiments illustrated that only detecting corresponging image point in target region can effectively reduce the number of non-corresponging image points, the time of corresponging image points detection and features matching, it also can effectively improve the speed of image registration, and finally complete image mosaic of tea plant.
关键词
图像拼接/图像排列顺序/相关性/矩形块/目标区域Key words
Image mosaic/Image sequence/Correlation/Rectangular-block/Target region引用本文复制引用
出版年
2016
国家科技期刊平台
NSTL
维普
万方数据