融合光谱与图像信息的河套蜜瓜糖度在线检测试验系统  

On-line Detection Testing System of Sugar Content for Hetao Melon Based on Spectra and Image Information

作　　者：高雄[1] 刘宇[1] 田海清[1] 陈亚莉[1] 韩宝生[1] 李哲[1] 

机构地区：[1]内蒙古农业大学机电工程学院,呼和浩特010018

出　　处：《农业机械学报》2015年第11期245-251,235,共8页Transactions of the Chinese Society for Agricultural Machinery

基　　金：国家自然科学基金资助项目(31160248);'十二五'国家科技支撑计划资助项目(2015BAD19B03)

摘　　要：针对蜜瓜糖度在线检测的需求,设计了融合漫透射光谱与图像信息的河套蜜瓜糖度在线检测试验系统,该系统包括硬件平台和软件系统两部分。硬件平台主要包括蜜瓜输送装置、光谱采集装置、图像采集装置和控制系统4部分。软件系统基于Microsoft Visual C++6.0语言,结合Omni Driver软件、Fly Capture2及Open CV软件开发。系统可实现蜜瓜光谱与图像信息的自动采集、显示及保存,可实现对漫透射光谱预处理,获取糖度检测所需光谱数据,对图像预处理,提取蜜瓜外观特征图像信息(R、G、B颜色值和蜜瓜体积)。在此基础上,系统可通过融合漫透射光谱与图像信息的蜜瓜糖度检测模型,计算蜜瓜糖度。系统同时可实现检测个数统计,外观特征信息及糖度的实时显示、保存等功能。测试试验表明,该试验系统检测1个样品用时1.2 s,糖度检测均方根误差为1.22,可满足河套蜜瓜糖度在线检测试验需求,为进一步开展河套蜜瓜糖度在线检测研究奠定了基础。Real-time detection of sugar content plays a crucial role in the estimation of Hetao melon quality. To reduce the effect of size on the diffuse transmittance spectra, an on-line detection testing system based on diffuse transmittance spectra and image information was designed for the evaluation of sugar content of melon. Two main parts （hardware and software） were involved in this device. The hardware includes delivery device, spectrum detection device, image capture device and control system. The software system was designed based on Microsoft Visual C ＋ ＋ 6.0 language and combined with the software of Omni Driver, FlyCapture2 and OpenCV. With the proposed system, firstly, the spectrum and images of melon were automatically obtained, displayed and saved. Secondly, the sensitive spectrum, color （R, G, B） and volume value of melon were derived by processing spectra and image information. Finally, the sugar content of melon was obtained according to the model of diffuse transmission spectrum and image information fusion. In conclusion, this device could automatically calculate the amount of melon, timely display and save the color value, volume value and sugar content of melon in computer, and detect a sample within 1.2 s with root mean square errors of 1.22 for sugar content. The proposed system was showed to be a promising tool for implementing on-line detection of Hetao melon sugar content.

关 键 词：蜜瓜 糖度 光谱 图像 在线检测 

分 类 号：S238[农业科学—农业机械化工程]