Abstract: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 this method is effective and practical to design an integrated control analysis software for a real-time system based on hyperspectral imaging technology.
