摘要
为进行昆虫图象的识别和分类 ,对昆虫图象几何形状特征的提取及测量进行了研究。提取了区域面积、边界周长、孔洞数、偏心率、形状参数、圆形性、似圆度、球状性和叶状性 9个直观易测的特征。采用判别分析方法对这些特征进行筛选得到 6个具有判别意义的特征 ,分别为区域面积、偏心率、形状参数、周长、似圆度、叶状性 ;剔除了与似圆度具有相关性的圆形性和球状性等特征。利用这 6个特征对 3种昆虫进行了识别 ,准确率均达到10 0 % ,表明本研究所提取的
Basic methods of feature measuring and extraction of 9 shape features such as region area, boundary perimeter, hole number, eccentricity, form factor, roundness, circularity, spherecity and lobation were introduced based on insect binary image. These features were measured on 3 insect species, i.e., Helicoverpa armigera Hübner with 50 observations, Ostrinia furnacalis Guenee with 40 observations, and Ectomocoris atrox St a ··l) with 55 observations. 6 determinant features, including area, eccentricity, formfactor, perimeter, roundness and lobation, were extracted using discriminant analysis method to produce discriminant functions. The recognition result using the discriminant functions to the 3 insect species showed that the accuracy is as high as 100%, which indicating that these features are very descriptive to the insect image, and very effective to the insect image recognition.
出处
《中国农业大学学报》
CAS
CSCD
北大核心
2003年第3期47-50,共4页
Journal of China Agricultural University
基金
国家自然科学基金资助项目 ( 3984 0 0 0 4)
国家高技术研究发展计划课题资助项目( 2 0 0 2AA2 4 30 31)
