Research And Implementation Of Voice Diaphragm On-line Detection And Control System

With the rapid development of technology, machine vision detection technology is increasingly being used in various production areas of the product testing industry.It promotes the corporations'production to become more automated, information and efficient.Using vision detection technology not only solves the traditional method of problems which it is labor-intensive,low efficiency,false detection and missed detection,but also improves the detection efficiency, reduces labor intensity and lower production costs while ensuring the product quality.The research object of this paper is the minitype voice diaphragm, its concentricity parameter is one of the important performance of the speaker sound quality. Based on anlysising the characteristics of voice diaphragm and combined with the design objective of the system, the thesis builds a online voice diaphragm detection system based on machine vision using computer and motion card. And high precision online detection of concentricity is achieved. According to its function,the system is divided into three subsystems,which is the vision subsystem, electrical and mechanical transmission subsystem and image processing subsystem. The main content of each subsystem can be summarized as follows:①to construct a optical imaging subsystem which can effectively reflect the inner circle and outer circle of the diaphragm.②researching image processing algorithms to quickly extract the inner circle and outer circle of voice diaphragm, and accurately calculating the concentricity parameters of voice diaphragm;③completing the development of detection systems, including automatic online detection, the control of removed electrical motion system and so on.In the course of product detection, the unqualified product need to be automatically removed from the line.The thesis design a air blown separation devices.And the Pneumatic control system's starting or stoping is control by DMC2410 card's general output ports. And the detail work principle is gived.After finishing the system's hardware design, the software of the machine vision system is detail discussion. The system of software uses multi-thread technology and each critical task to take a thread to take advantage of system resources and improve the efficiency of the system. According to the voice diaphragm's features,the system extractes the inside and outside circle parameters separately. First, using threshold segmentation and tracking extracts the outer circle parameters,and then according to the circular cylindrical parameter determines the local area, and using the improved random hough transform algorithm to extract the inner circle's parameters. Finally,according to the inner circle and outer circle's parameters, using least squares algorithm to fit the outer circle and inner circle and calculate the concentricity parameter.Through a large number of experimental measurements, the system has a good reliability and can effectively eliminate unqualified voice diaphragm.