摘要
群体水果动态图像的采集系统是水果品质实时检测与分级机器人系统的关键组成部分之一,主要由计算机、摄像机、图像采集卡、涡流式接近开关、水果和动态图像采集软件等组成.涡流式接近开关作为水果运动状况检测传感器,用于检测水果位置和触发动态图像采集.群体水果动态图像采集软件在Vi-sualC++6.0环境下结合MIL-LITE函数库实现.根据水果实时自动检测和分级的特点及功能要求,提出了群体水果动态图像的时间序列连续场扫描采集方法,在来自于水果运动状况检测传感器的外触发信号的触发下,计算机视觉系统采集从时间t1到时间tn期间视场内连续运动的水果图像,这些渐进的时间序列水果图像在计算机内被综合后即可检测视场内的每一个水果的外观品质指标,用于分级控制.试验结果表明,该系统可在外界输入信号的触发下,以每秒12帧的速率实时采集群体水果的动态图像,研究结果达到了预期的目标.
The real-time image acquisition system of massive numbers of moving fruit is one of the key sub-systems of quality inspection and grading robot of fruits, which consisted of a computer, a Panasonic CP410 color camera, a Meteor Standard/Ⅱ grabber, eddy proximity sensors and a set of real-time image acquisition software. The proximity sensors were used to track the fast moving and rotating fruits on the conveyers and then trigger the camera and the grabber to acquire image of moving fruits, and the software was programmed with Microsoft Visual C++6.0 and MIL-LITE function library. In the light of the characteristics and requirement of real-time inspection and grading, the continuous area-scans approach was put forward to acquire the image of fruits from time t1 to time tn in a time-series manner. These progressive time-series digital images were synthesized in system memory so that the overall composite sides of the fruits could be examined. This new approach enables the large coverage of multiple fruits simultaneously and utilizes less equipment for an entire machine. The test results verified that this system could grab 12 frames per second with the triggering from the proximity sensors.
出处
《浙江大学学报(农业与生命科学版)》
CAS
CSCD
北大核心
2004年第2期147-152,共6页
Journal of Zhejiang University:Agriculture and Life Sciences
基金
国家自然科学基金资助项目(30270763)
国家"863"计划项目(2001AA422230).