| 本文已被：浏览 958次 下载 734次 |
码上扫一扫！ |
| 计算机图像识别技术在小麦种子精选中的应用 |
|
贾佳1, 王建华1, 谢宗铭2, 杨丽明3, 孙宝启1,4, 孙群1
|
|
|
| (1.中国农业大学 农学与生物技术学院/北京市作物遗传改良重点实验室, 北京 100193;2.新疆农垦科学院 分子农业技术育种中心, 新疆 石河子 832000;3.中国农业大学 理学院, 北京 100083;4.北京市农林科学院, 北京 100097) |
|
| 摘要: |
| 以农大399小麦种子为试验材料，通过图像识别技术获取单粒小麦种子的形态物理指标，拟通过研究这些形态物理参数与小麦种子活力的相关性，为图像识别技术在作物种子加工工艺和参数精确选择提供参考。首先采用平板扫描仪获取单粒小麦种子的PNG图像，采用本实验室自主开发的种子形态自动化识别软件自动提取单粒种子(400粒)的RGB、Lab、HSB、长度、宽度和投影面积等形态物理指标，然后进行垂直玻璃板发芽试验，以发芽第5天的幼苗长度、鲜重和简易活力指数作为种子质量指标，相关分析结果表明：农大399小麦种子的幼苗鲜重与种子投影面积、长度和宽度的相关系数分别为0.45、0.40和0.37，均达到显著相关。将种子按投影面积、长度和宽度进行分组，实验结果表明幼苗鲜重和简易活力指数随着种子投影面积、宽度和长度的增加而增加，且达到极显著差异水平，生产上可以根据对种子质量或数量的需求灵活设置相关加工参数。 |
| 关键词: 小麦 种子形态自动化识别软件 种子形态物理特性 种子精选 种子活力 |
| DOI：10.11841/j.issn.1007-4333.2014.05.25 |
| 投稿时间：2014-01-07 |
| 基金项目:公益性行业（农业）科研专项项目（201303002）；新疆生产建设兵团“十二五”项目（2012BB046） |
|
| Wheat seeds selection based on computer image recognition technique |
|
JIA Jia1, WANG Jian-hua1, XIE Zong-ming2, YANG Li-ming3, SUN Bao-qi1,4, SUN Qun1
|
| (1.College of Agronomy and Biotechnology/Beijing Key Laboratory of Crop Genetics Improvement, China Agricultural University, Beijing 100193, China;2.Center for Molecular Agro-biotechnology and Breeding, Xinjiang Academy of Agricultural andReclamation Sciences, Shihezi 832000, China;3.College of Science, China Agricultural University, Beijing 100083, China;4.Beijing Academy of Agriculture and Forestry Sciences, Beijing 100097, China) |
| Abstract: |
| In this study, the correlation between the physical traits and seed vigor of wheat seed (Nongda 399) was analyzed, several morphological and physical traits was captured with image recognition technology.The objective of this research was to test the feasibility on selection of wheat seeds processing technology and the parameter of Seed Identification software developed by our lab.The PNG format images of 400 wheat seeds were acquired with flatbed scanner, and the color features of wheat seed such as RGB, HSB, Lab, gray scale, and width, length and projected area were extracted automatically and quickly with the software.The Seed Identification software could identify the image and record related seed physical information, and then output all the information into an Excel file automatically.The identifying results could be achieved in 1 second, very quickly, with errors lower than 2%.Seedling length, fresh weight after 5－day germination and simple vigor index were used as the indicators of seed quality.The results of correlation analysis showed that projected area, length and width were all significantly correlated with seedling's fresh weight.The correlation coefficients were 0.45, 0.40 and 0.37 respectively.Wheat seeds were grouped into three intervals according to projected area、length and width, verification test results showed that seed vigor increased along with the increase of seed projection area、length and width.There was significant difference among different groups.Therefore, according to the market demands for seed quality or quantity, people could determine the optimum seed selection technology and parameters and get the high vigor seeds from a seed batch quickly and accurately. |
| Key words: wheat seed Identification seed morphological and physical traits seed selection seed vigor |
