The Research of Image Collection Method for Sediment Online-Detection

The water level and flow of middle and small rivers has been achieved online-monitoring. However, the sediment detection is still in the manual observation. We propose the sediment online-detection system of in flow-water based on image analysis. We capture the underwater images by special camera device, and then transmit images to monitoring centers through channel. Furthermore, we implement special video capture card to preprocess, segment, identify the hydrology image and calculate the sediment of sample. The characteristic of system is image segmentation algorithm based on a wavelet level Markov model, the multi-direction one-dimensional wavelet transformation microscopic image denoising algorithm. We build the preliminary model based on the feature analysis of image. The experiment shows that sediment detection system could get the sediment hydrological of middle and small rivers online and provides a theoretical basis for the sediment detection.