|
|||||
|
|
| 基于多分辨率分析的储粮害虫图像预处理研究 | |
| 引用本文： | 胡玉霞,张红涛,毛罕平,韩绿化.基于多分辨率分析的储粮害虫图像预处理研究[J].农机化研究,2012,34(8):160-163. |
| 作者姓名： | 胡玉霞 张红涛 毛罕平 韩绿化 |
| 作者单位： | 1. 郑州大学电气工程学院,郑州,450001 2. 华北水利水电学院电力学院,郑州,450011 3. 江苏大学现代农业装备与技术省部共建教育部重点实验室/江苏省重点实验室,江苏镇江,212013 |
| 基金项目： | 国家自然科学基金，河南省高等学校青年骨干教师资助计划项目，华北水利水电学院高层次人才科研启动项目 |
| 摘 要： | 粮虫种类多、体形小且形态结构比较复杂,须借助于高分辨率的图像进行准确的细分。针对高分辨率的粮虫图像,提出基于多分辨率图像分析的粮虫图像预处理方法。本算法首先降低图像的分辨率,并对低分辨率的图像滤波,然后将恢复为原分辨率后的图像与原图像进行"与"运算,并进一步滤波。该法有效去除了对粮虫识别贡献率较小的腿和触角等部分,保留了对识别分类贡献率较大的头部和鞘翅等部分,且预处理后的目标边界比较平滑,同时使算法运行的效率提高了15.12倍,并通过实验进一步验证了该方法的可行性。 |
| 关 键 词： | 储粮害虫 多分辨率分析 图像预处理 形态学滤波 特征提取 |
Image Preprocessing of the Stored-grain Insects Based on Multi-resolution Analysis |
|
| Hu Yuxia , Zhang Hongtao , Mao Hanping , Han Lvhua.Image Preprocessing of the Stored-grain Insects Based on Multi-resolution Analysis[J].Journal of Agricultural Mechanization Research,2012,34(8):160-163. | |
| Authors: | Hu Yuxia Zhang Hongtao Mao Hanping Han Lvhua |
| Institution: | 1.College of Electric Engineering,Zhengzhou University,Zhengzhou 450001,China;2.Institute of Electric Power,North China Institute of Water Conservancy and Hydroelectric Power,Zhengzhou 450011,China;3.Key Laboratory of Modern Agricultural Equipment and Technology,Ministry of Education & Jiangsu Province,Jiangsu University,Zhenjiang 212013,China) |
| Abstract: | The stored-grain insects were of many species,small size and complex morphology.The high resolution images of the stored-grain insects were needed to acquire to be accurate identification.The image preprocessing method based on multi-resolution analysis was proposed for the high resolution images of the stored-grain insects.The algorithm first reduced the resolution of the image and filtered the low-resolution image.Then the "and" operation was conducted between the image restored to the original resolution and the raw-resolution image.Lastly the image was further filtered.The algorithm effectively removed the leg,antennae and other parts that had less contribution to the recognition of the stored-grain insects,and effectively retained the head,elytra and other parts that had larger contribution.The boundary of the object was smooth by the pretreatment,and the algorithm efficiency improved15.12 times.The experiment showed that it was practical and feasible. |
| Keywords: | stored-grain insects multi-resolution analysis image preprocessing morphological filtering feature extraction |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |