当前位置: 首页 > 文章 > 农产品线扫描高光谱成像系统的集成标定方法 农业工程学报 2012,28 (14) 244-249
Position: Home > Articles > Integrated calibration of line-scan high spectral imaging system for agricultural products Transactions of the Chinese Society of Agricultural Engineering 2012,28 (14) 244-249

农产品线扫描高光谱成像系统的集成标定方法

作  者:
王海华;李长缨;梅树立;李民赞
单  位:
佐治亚大学生物与农业工程系;中国农业大学现代精细农业系统集成研究教育部重点实验室
关键词:
扫描;标定;变形;高光谱成像;集成
摘  要:
高光谱成像技术在农业和食品检测中的应用日益广泛,线扫描高光谱成像系统的标定是获取高精度图像的基础。该文从系统集成部件的调整顺序、扫描速度控制、导轨偏移校正、光谱图像变形标定和光谱波长偏移校正等几方面系统地讨论了一种线扫描高光谱成像系统的调整和标定。通过对棋盘格图案的扫描,校正光谱图像在CCD上成像的位置,给出了导轨偏移校正及速度控制的具体计算公式。研究了光谱图像水平(Smile)和垂直(Keystone)光学系统变形,并标定了图像2个维度变形误差,分别为1.26和0.04mm。对照笔型氪气校正灯标准光谱,给出了光谱波长移位的二次校正方程。实践证明该文提出的方法能快速、有效地完成图像采集试验的设备准备工作,为相近研究领域的科研人员提供参考和帮助。
译  名:
Integrated calibration of line-scan high spectral imaging system for agricultural products
作  者:
Wang Haihua 1,Li Changying 2,Mei Shuli 1,Li Minzan 1(1.Key Laboratory on Modern Precision Agriculture System Integration Research of MOE,China Agricultural University,Beijing 100083,China;2.University of Georgia,Department of Biological and Agricultural Engineering,Georgia 31793,USA)
关键词:
scanning,calibration,deformation,hyperspectral imaging,integration
摘  要:
Hyperspectral imaging(HSI) technology has been used widely in agriculture and food industry.The paper discussed the integration and calibration methods for a line-scan HSI system,such as the sequence of the system adjustment,scanning speed and position control of the linear slide,spectral accuracy calibration,and spatial distortion calibration of the spectral image.Detailed methods and steps were provided to ensure the high fidelity of the image.The "smile"and "keystone"distortions of the spectral images were also taken into account.The calibration results showed that the spectral and spatial errors were 1.26 nm and 0.03 mm,respectively.A quadratic equation was used to correct the wavelength positions based on the standard spectrum of a Krypton lamp.It is proved that this method provided preferences to prepare image acquisition system quickly and efficiently.

相似文章

计量
文章访问数: 9
HTML全文浏览量: 0
PDF下载量: 0

所属期刊

推荐期刊