项目名称: 基于外观信息感知的脱绒棉种内部品质检测机理研究
项目编号: No.31260290
项目类型: 地区科学基金项目
立项/批准年度: 2013
项目学科: 农业科学
项目作者: 李景彬
作者单位: 石河子大学
项目金额: 53万元
中文摘要: 种子内部品质的快速无损检测对育种和农业生产都有重要的意义。研究表明,种子的内部品质与外观特征信息存在密切的关系。 本项目以脱绒棉种为对象,基于外观特征信息研究其内部品质的检测机理和方法,构建脱绒棉种内部品质检测系统。该研究从脱绒棉种的样本采集入手,首先基于机器视觉技术,开发相应的图像处理算法,从脱绒棉种的彩色图像中提取颜色、纹理、合点、脊、珠孔等外观特征信息;而后通过传统活力检测方法,测定反映棉种内部品质的发芽率、活力指数、TTC脱氢酶、ATP含量等生理指标,明确反映种子内部品质的主要评价指标;运用正交试验设计方法,确定显著影响棉种内部品质的外观特征参数组合;基于多元回归分析和模糊综合评判算法,建立外观特征信息与反映种子内部品质的主要评价指标的之间关系的预测评价模型,并进行实验验证,最终构建内部品质快速检测系统。项目完成后,可为实现脱绒棉种按照活力在线检测分级奠定基础。
中文关键词: 脱绒棉种;内部品质;活力检测;外观特征;高光谱图像技术
英文摘要: Rapid nondestructive testing of internal quality of seed has an important significance on breeding and agricultural production. The internal quality has a close relationship with the appearance information of seed. The purpose of this project is to study on the detection mechanism of the internal quality and build the detection platform on the basis of appearance information. Firstly, collect the sample of cottonseeds and extract appearance information, such as color, texture, chalaza, ridge, micropyle. Secondly, measure the physiological indicators, such as germination rate, vigor index, TTC-dehydrogenase and ATP content. Thirdly, determine the appearance parameters which affect the internal quality significantly with orthogonal design. Finally, on the basis of multiple regression analysis and fuzzy comprehensive evaluation, establish the evaluation model of the relationship between the internal quality and the appearance information, then build the system for detecting internal quality of cottonseed. Also,it can lay the foundation for online cottonseed inspecting and grading by using the vigor index.
英文关键词: Delinted cottonseed;Internal quality;vigor detection;Appearance characteristics;hyperspectral imaing technique