Research On Embedded System Of Geometric Parameters Measurement Of Dropping Pill

With the mass production of dropping pills in food,tobacco,Chinese medicine and other industries,the producers put forward higher requirements on the geometric parameters measurement of dropping pill.The producers have used manual detection method,which is not only time-consuming and laborious,but also cannot guarantee the reliability and accuracy of the data.Therefore,the study of the system for detecting the geometric parameters of soft dropping pill has important practical significance.The detection scheme for measuring the geometric parameters of soft dropping pill is designed by analyzing the actual demand of the detection for dropping pill.The system captures the image of dropping pill by using ARM-Linux embedded system and extracts the pill from the digital image,then detects the geometric parameters of dropping pill by fitting the edge curve.By analyzing the characteristics of the image and the gray histogram of dropping pill,the image processing algorithms for the image of dropping pill,such as image enhancement,image de-noising,image segmentation and object extraction are studied when running in embedded system environment.These algorithms are implemented by using C++ programming language.By the comparison and analysis of the experimental results,the most suitable image processing algorithms will be found.Based on the MVC architecture model,the software architecture of the geometric parameters detection program of the dropping pill was designed.The design and implementation of image acquisition,system calibration and image measurement are also carried out.Finally,based on the analysis of the experimental data,the results show that the proposed image measurement algorithms can meet the design requirements and has good stability.