摘要
提出利用边缘检测及角点提取来识别蝶类昆虫的方法。通过用四种边缘检测算子对鳞翅目蝴蝶进行边缘检测并作比较分析,从中找出鳞翅目昆虫边缘检测的最佳边缘检测算子,然后对边缘检测后的图像进行了角点提取特征信息,通过实验能够得到蝶类图像的特征,为下一步蝶类识别奠定了基础。
In this paper,the methods of edge detection and corner extraction are presented for the recognition of rhopalocera insects.Four kinds of edge detectors were used to detect image edge of the lepidoptera butterflies and the detective results were analytically compared,among them we found the best edge detector which can detect the edges of lepidoptera insects,then corner extraction for the characteristic information of the edge-detected image was done.The characteristics of the image of rhopalocera are able to collect through the experiment, whichprovides a basis for the rhopalocera recognition in next step.
出处
《计算机应用与软件》
CSCD
2009年第10期78-80,共3页
Computer Applications and Software
基金
山东省教育厅基金项目(J03P52)
关键词
鳞翅目昆虫
边缘检测
轮廓提取
角点检测
Lepidoptera insects Edge detection Contour extraction Corner detection
