摘要
研究了非结构环境复杂背景下图像拼接方法,实现了温室环境下红掌图像的拼接。以温室自然光照环境下随机垄地为研究对象,提出先采用2G-R-B算法提取红掌、再采用SIFT算法进行图像特征匹配的方式,并对使用SIFT算法的图像拼接算法进行了改进,大大降低了原拼接算法在高噪声环境下的运算量及处理时间,实现了高噪声环境下红掌图像快速有效的拼接。在智能对靶施药系统的基础上进行实验,结果表明,所提出的方法拼接速度为0.083 s,误判率为7.3%。
An image stitching method in the complex environment of non-structural context was studied. The anthurium image mosaicing under greenhouse environment was achieved. A random ridge by natural light environment in greenhouse was researched. An anthurium image stitching method based on SIFT only was improved by using a method that extracting anthurium by a 2G - R - B method firstly, and registering image by SIFT method secondly. The proposed method can greatly reduce the computation and processing time under high-noise environment. And a quickly and efficiently anthurium extraction method under high-noise environment was achieved. Experiments were carried on intelligence target spaying system. The results showed the mosaic time-consuming was 0. 083 s and false positive rate was 7.3%.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2013年第4期223-227,共5页
Transactions of the Chinese Society for Agricultural Machinery
基金
国家高技术研究发展计划(863计划)资助项目(SS2012AA041507)
中国博士后科学基金资助项目(2011M500450)
中央高校基本科研业务费专项基金资助项目(2012QJ119)
中国农业大学基本科研业务费研究生科研创新专项资助项目(2012YJ106)
关键词
红掌图像
温室
自动拼接
SIFT算法
特征匹配
Anthurium image Greenhouse Automatic mosaicing SIFT algorithm Feature matching
