Discrimination of the imaging quality based on the sharpness and illuminance factor

In the visual inspection with the machine system, the prerequisite for the detection success is to get a clear and stable image with high signal to noise ratio. For the current lack of quality assessment on the collected images, an experimental evaluation method is proposed for the quality of the collected image based on the image sharpness parameters and the illumination factor of the CCD target surface. Heavy rail surface was selected as the research object, and a flag was set as an object. The first thing was to get the best image sharpness point of the object, followed by employing wavelet parameters and the least squares method (LS-SVM) to identify the object. Meanwhile, the loss rate of the object size was analyzed to get the best illumination factor of the CCD target surface, and thus obtaining the best spots for global capture of image quality. This method gave a more comprehensive consideration on various factors on CCD imaging, and proven to be quite practical.