Ultraviolet（UV）image technology is an effective method to detect the discharge of high-voltage equipment. However

UV spots are sometimes not clear in actual practice making it hard to distinguish them from noise spots. To resolve this problem

we propose an image superposition method（ISM）to analyze UV images. We count the occurrence number of UV spots in each pixel spot of images taken in the same discharge region and generate a two-dimensional matrix where each element represents the frequency of UV spots occurred in the pixel spot of UV images. Since UV spots appear more frequently in the discharge region

in the matrix

elements within the discharge region have larger values. We can separate the discharge region in the UV images by setting a proper threshold. Finally

grading ring defect discharge test is designed

the relationship between the discharge area and the observation distance is obtained

as well as it is proved that the improved image processing method is reliable and effective

and having a better capacity of dealing with the ultraviolet image

which formed by corona discharge with unapparent light spot.