The Water level detection algorithm using the accumulated histogram with band pass filter

초록

In this paper, we propose the robust water level detection method based on the accumulated histogram under small changed image which is acquired from water level surveillance camera. In general surveillance system, this is detecting and recognizing invasion from searching area which is in big change on the sequential images. However, in case of a water leel detection system, these general surveillance techniques are not suitable due to small change on the water surface. Therefore the algorithm introduces the accumulated histogram which is emphasizing change of water surface in sequential images. Accumulate histogram is based on the current image frame. The histogram is cumulating differences between previous images and current image. but, these differences are also appeared in the land region. The band pass filter is able to remove noise in the accumulated histogram. Finally, this algorithm clearly separates from the water level value on the image space to the real water level on the real space using calibration table. The detected water level is sent to the host computer with current image. To evaluate the proposed algorithm, we use test images from various situations.

키워드