Research Of Online Video Detection Methods For A Bottle Cap On The Production Line

This paper is to study online video detection methods for bottle caps on the production line. Based on completing the design of the detection device, the paper mainly realizes the online video detection for a bottle cap. Among them, the detection includes the color difference detection and the integrity testing of bottle cap. This paper realizes the online video detection for a bottle cap on the production line through the research and design of the following links.First of all, the paper completes the hardware and software design of the detecting device. According to the characteristics and needs of bottle cap detection, the paper selects the appropriate hardware devices, designs and implements the online video detection device for a bottle cap. It not only uses CMOS OV7670 as the video module, but also uses FPGA as the control unit. With analyzing the principle of the various parts of the hardware device, in the Quartus II software programming environment, the device realizes the menu-parameters settings of the external VGA monitor by the Verilog HDL programming language.Secondly, the device performs the color difference detection of bottle cap. According to the characteristics and needs of bottle cap detection, this paper first sets the template image, and divides the window area of the template image which was detected. In order to obtaining the difference value of each window area, through the Verilog HDL programming language, the device makes use of color spaces and color difference formulas. And then the device identifies the unqualified bottle caps, so that it realizes the color difference detection of bottle cap.Last but not least, the device performs the integrity detection of bottle cap. The device first uses CMOS camera to read and store the video information of the detected bottle image, and completes the HDL transplantation of algorithm procedures what combined with the working principle of edge detection algorithm. So the device achieves the edge detection of the online video images. Noise was reduced by improving the detection algorithm, and the device got a clear and coherent edge image. Afterwards,setting the template based on image difference principle, and using simple or, XOR between images, the device designs and completes the integrity test of cap bottle. And through the analysis of the whole integrity detection algorithm, the device improvesdetection algorithm again, what makes the device under the premise of ensuring the detection quality. In this way, the design simplifies the operation process, and also reduces the amount of computation.On the whole, the design achieves the expected results, achieves the design of detection device, and completes both the color difference detection and the integrity detection of bottle cap. The device is high precision, fast response, automatable and intelligent. It not only has the good real time performance, adaptability and portability, but also has zero damage to human body. It has a broad application prospect because it can be widely applied in poor environment for long time.