Study Of Quality Inspection System Of The Liquid Mosquito-repellent Incense In Bottle Based On Machine Vision

Machine vision system use machines to replace the human eye to measure and determine the relevant object features. Machine vision technology with high speed and high resolution and high precision can test the object of detection with no-direct contact with it. Therefore, it is widely used in the processes of industrial automation products to detect the quality of products and automatically to control the progress of product automatically.Currently the major manufacturers of liquid mosquito-repellent incense mainly use artificial visual methods to detect the mosquito coils, liquid level, and crack and other quality parameters after being filled with liquid Therefore; machine vision quality inspection system has broad prospects and needs. In the time of a mosquito-repellent incense coil plant research we found that the liquid mosquito-repellent incense coils manufacturers mainly relied on product quality testing and the use of artificial visual Graduated scale and balance to measure the instrument sampling. Detection of such defects is slow, inefficient, impossible to detect all the products accurately, and to increase the labor cost. Liquid mosquito-repellent incense coils based on machine vision quality inspection system can meet the need of high speed, high precision, high and mass production of quality on automated testing.In this paper, combined with the actual characteristics of the liquid mosquito-repellent incense products and actual demand of the factory; it proposed the line of liquid mosquito-repellent incense Quality Inspection System Design program. And it also made in-depth analysis and research on hardware, image processing and the software design. In allusion to the characteristics of the product, it succeeded in choosing the camera, light source, lens and other key hard wares. Through image processing, it respectively validated the effects of different algorithms in mosquito- repellent incense coils. The appropriate system was chosen by comparing different algorithms. Product Image edge detection applied Contourlet transform-based detection methods, and on the surface feature extraction and recognition, it applied the calculation based on the continuity of thought. Based on the VC++ 2005, it designed the human-machine interface and completed the systemic software design.Theoretical analysis and experimental results demonstrated that the liquid mosquito-repellent incense detection system is reasonable, feasible, and the effectiveness of key quality parameters on the bottle size and the net content testing. It also solved most issues of the actual testing, and established a good foundation for further research and development. Quality testing in the application of liquid mosquito-repellent incense has some practical value.