Phosphor Magnet Powder Flaw Detection System Of The Wheel Band Surface

The wheel band surface phosphor magnet powder flaw detection system is a new generation of industrial online flaw detection system, which has images collection based on darkroom conditions, image processing, automatic detection of the flaw. The innovations of this system are the real-time online processing of CCD collecting images and real-time alerting of suspicious image, which provides examples for the flaw detection of other metal surface. ( steel, artillery guns, etc.)Based on CCD imaging, the operation principle and structure of the phosphor magnet powder flaw detection system of a wheel band surface and the display system of computer image have been introduced in detail in this paper. The UV- spectrum, the excited spectrum of the phosphor magnet powder and the working waveband of CCD camera have been analyzed in this system, the type of ultraviolet source, CCD models, work-band and parameters are determined. This system uses the 365nm ultraviolet source of the largest spectral power system used for the 365nm and high-level testing pixels, high sensitivity CCD camera. In order to stop the visible light to shine the wheel band, we add a glass filter in front of the ultraviolet source, which can filter effectively smaller than the 300nm harmful ultraviolet ray and bigger than 420nm the useless visible light, also has the high transmissibility to the spectrum power more centralized 365nm ultraviolet light. It enhances effectively overall system precision and sensitivity. In order to reduce the influence of the pseudo-flaw in the image, such as the water drop, slash, surface pollution and so on, and eliminate the light source and the CCD camera vibration influence, we have designed a set of image real-time processing and the examination software through researching and analyzing the image characteristic on the processing aspect of image. First, this software must carry on smooth processing to the image, eliminate effectively noise and pseudo-flaw disturbance. Smooth processing causes the entire image ambiguity at the same time to eliminate noise and pseudo-flaw, then we withdraw the edge with the derivative.