基于图像传感器的片烟规格测量方法

针对传统工艺对打叶后片烟面积的检测和分类方法的缺点，提出了基于图像传感器的片烟规格测量方法。由CMOS图像传感器采集工作台上的片烟图像信息，并以统一的亮度阈值对图像进行二值化处理；然后采用硬件方法提取二值图像的边界信息，再根据连续区段标记算法统计出各个连续区域的面积，进而最终得到各个面积段片烟的比例。实验证明该测量方法准确可靠，测量结果相对误差小于2％；实时性好，为打叶工序工艺参数的快速调整提供了依据。

Measurement of strips size based on image sensor

For the shortcomings of the traditional processes of strips area measurment and classification, a method for strips size measurement based on image sensor was put forward. The image information of the strips on the working platform was scanned by CMOS image sensor and binary processing is implemented under an uniform threshold value directly. Then the edge information of the binary image was extracted by hardware method. Area of each successive sections were computed by connected component labeling algorithm. Finally the proportions of every span of strips area were obtained. The experiment result proves that the method is accurate and reliable. The relative error of the result is less than 2%. Meanwhile it has a good realtime performance, which provides a reference for fast adjustment of the threshing process parameters.