昆虫图像分割方法及其应用

Methods of insect image segmentation and their application

昆虫图像自动鉴定是一种快速鉴定昆虫的方法,图像分割则是其中关键步骤。通过搜集和整理国内外近年来针对昆虫图像的分割方法和研究,发现对昆虫图像分割的研究日趋增多。随着计算机图像技术的发展,昆虫图像分割方法吸收了许多图像分割领域中新兴的方法,诸如采用水平集、边缘流以及结合形状、纹理、色彩等多种要素的智能分割(如JSEG方法)等。虽然大量的图像分割方法被引入到昆虫图像研究中,但是目前分割技术依然是阻碍昆虫图像广泛应用的关键。本文经过总结和分析,发现目前昆虫图像分割研究的往往在各自的测试集上有良好表现,但是缺乏统一的评价标准,因此很多方法在昆虫图像中应用难以推广。针对研究中的存在的这些问题,需建立良好的昆虫图像分割评价体系,本文建议通过建立统一的昆虫图像库以及对昆虫图像分割的评价方法深入研究,并且这些工作是当前昆虫图像分割研究亟待完善任务。

Automatic identification on insect images is a rapid identification method,and image segmentation is a key process in insect image automatic identification.Based on searching and sorting the recent references on insect image segmentation,we found that the researches on insect image segmentation are increasing rapidly.With the development of image processing technology,all kinds of segmentation methods such as level set,edge-flow and JSEG（a segmentation method integrating shape,color and texture information of images）had been tested in insect image segmentation.Although these methods performed well on their own test sets,there is no sharing image dataset or rules to evaluate these methods.It is necessary to build a system to evaluate the increasing insect image segmentation methods.In this review,we suggest creating some standard datasets and doing researches on evaluation of segmentation methods which are desiderated in insect image segmentation.