农产品品质检测系统的高光谱成像控制软件设计  

Control and Analysis Software Design of Hyperspectral Imaging System for Detection in Agricultural Food Quality

作　　者：赵娟[1] 彭彦昆[1] 郭辉[1,2] 陶斐斐[1] 张雷蕾[1] 

机构地区：[1]中国农业大学国家农产品加工技术装备研发分中心,北京100083 [2]新疆农业大学机械交通学院,乌鲁木齐830052

出　　处：《农业机械学报》2014年第9期210-215,共6页Transactions of the Chinese Society for Agricultural Machinery

基　　金：公益性行业(农业)科研专项经费资助项目(201003008)

摘　　要：利用VS2010与Matlab混合编程方法设计了用于农产品品质指标检测的高光谱成像在线检测系统的控制分析软件,包括仪器参数设置模块、信号检测与控制模块和数据采集与分析模块,完成了图像采集、图像合成、运动控制、数据提取分析及存储、显示功能。该控制分析软件设计提高了高光谱成像技术应用的实用化,实现了对农产品品质指标的无损、实时、快速检测分析。Hyperspectral imaging technology was applied for quality detection of agricultural product. The research was based on laboratory setup and application. A software system was developed for real-time, rapid detection of agricultural product based on hyperspectral imaging technology. A real-time hardware system consisted of sample feeding unit, on-line detection unit and sample separation unit. The important on-line detection unit comprised charged couple device （CCD） camera, imaging spectrometer, lens, data acquisition card, light source. Sensor was developed to acquire image of intact samples. Control and analysis software was designed to detect quality parameter in agricultural food based on the VS2010 and Matlab platforms. The software designed included instrument parameter setting module, signal detection and control module and data collection and analysis module. Each functional module was divided into several sub function modules according to different function. These three main modules were integrated together to achieve the software functions of images acquisition, images synthesis, motion control, data extraction, storage and display. For data analysis, the study emphased on analyzing the process of image synthesis in the interface of image display step by step. At the same time, for efficient image processing and spectral information in the software system, image data correction and choice of region of interest is required to display and calculate data in the whole software. Because of hyperspectral imaging technique included image information and spectral information, two part information for detecting samples was represented different preprocessing analysis method to eliminate noise and correct the original information. This Study makes use of all above mentioned hardware and method to design practical analysis software to detect a variety of quality parameters of meet. It describes design ideas of the meet detection flowchart and shows the graphical user interface （GUI）. The result stated that

关 键 词：农产品 高光谱成像 品质检测 控制分析软件 

分 类 号：O657.3[理学—分析化学] TP274.5[理学—化学]