摘要
全景图能够提升旅游体验,是智慧旅游中重要的内容之一.文章改进了Harris算法中的梯度算子和角点响应函数,以减缓角点提取时的聚簇现象,降低角点冗余;改进了RANSAC算法匹配特征点的选取方式,降低了算法的运行时间.结合这两种改进算法,生成了长江黄金邮轮重要部位的全景图,提升了旅游服务智慧,满足了智慧旅游的应用要求.实验表明:两种改进算法具有特征点提取均匀、配准速度快的优点.
Panorama is one of the important content of intelligent tourism which can improve tourism experi- ence. The gradient operator and corner response function of Harris algorithm were improved so that the corner points clustering could be decreased, also the redundancy of comer points decreased. The selecting method of feature point of RANSAC algorithm was improved so that the time cost could be decreased. Combining with two improved algorithms, important parts of Yangtze gold cruises＇ panoramic images had been generated so that the tourism could touch the service intelligence and the application requirements of intelligent tourism could be met. Experiments showed that the two improved algorithms＇ advantages were uniform feature point extraction and high speed of registration.
出处
《重庆文理学院学报（社会科学版）》
2013年第5期101-107,共7页
Journal of Chongqing University of Arts and Sciences（Social Sciences Edition)
基金
重庆市教委科研项目(KJ120611)
重庆师范大学横向课题(60102000026)
关键词
全景图
图像配准
智慧旅游
长江黄金邮轮
panorama
image registration
intelligent tourism
Yangtze gold cruises
