摘要
以无屏闪光照系统作为光源,物料通过Y-Roller均散装置摊开后高速抛出时,CCD摄像机 采集到物料图像并送入PC机;在光学滤色镜组的作用下,烟草与异物的色度值拉开,以便后续的识别与剔除处理;建立中心阈值库以加快判别阈值的确定,用浮动阈值提高判别阈值的机敏度,提高系统的识别率。系统原理样机对白纸、鸡毛等典型异物的识别率可达到60%以上。
With Flash illumination system without screen as light source, when materials pass through Y-Roller uniform spreading unit and being ejected, CCD camera samples the images of materials and sends them into PC. With the action of optical filter, the chromatic difference between tobacco and foreign matter is enhanced, thus is convenient for the follow-up recognition and elimination. The establishment of central threshold base speeds up the judgment and determination of threshold and the sensitivity of judgment threshold is improved by floating threshold, thus enhances the recognition ratio of the system. The recognition ratio of the prototype for the typical foreign materials such as white paper, chicken feather, etc. can be more than 60%.
出处
《光电工程》
CAS
CSCD
北大核心
2003年第5期51-54,72,共5页
Opto-Electronic Engineering
基金
国家计委基金资助项目 (2000-1882)。
关键词
实时测量
图像处理
CCD摄像机
浮动阈值
烟草
异物
Real-time measurement
Image processing
CCD cameras
Float threshold
Tobacco
Foreign matter
