The efficiency improvement of image-based snow-cover measurement system using histogram projection and thinning method

In this paper, a snow-cover measurement algorithm was developed for improving the snow-depth analysis on the Internet-based visual snow-cover monitoring and measuring system (Cliview-High). Cliview-High which analyzes the visual image of drifted snow and measures the amount of snowing is consisted of a reference ruler, a CCD camera, an industrial PC with image processing unit, and an image data DB server connected by Internet. The snow-depth is estimated based on the position of the lowest snow-uncovered marker on the ruler which is auto-detected from the captured image. The distortion in the captured image is calibrated by the Newton´s polynomial function for the accurate measurement. The step for each image to be processed is, first, brightness and contrast improvement as preprocessing. Second, a feature point of the ruler is periodically detected in specified image region. Then the proposed algorithm is performed following by a labeling algorithm after ROI(region of interest) cropping which includes a reference ruler. Finally measured data is filtered by QC(quality control) technique for reliability.