Depth Estimation Methods For Determining Coal Level In Coal Bins

Under the poor imaging environment of high dust, high humidity, low illuminate, this paper proposes a measurement method of depth estimation based on the controlled laser power and image recognition technology. The corresponding prototype models are developed. The depth values of the coal level in coal silo are determined by compute the transverse offsets of spot points and image center points, query pre-measured calibration table corresponding to different offsets. It is proposed that the fast denoising algorithm based on Poisson Uunbiased Risk Estimate-Linear Expansion of Thresholds (PURE-LET) in the wavelet domain of laser spot on the surface of coal level. An unbiased estimate PURE of wavelet coefficients estimate MSE in the Poisson noise is given, and the wavelet coefficients estimates are written a set of basic linear combination of the threshold functions to improve the speed of algorithm. The denoising results of simulated images and real laser spot images show that this algorithm has better suppress ability to Poisson noise than the other three typical image denoising algorithm, which has the advantage of fast calculation. Linear camera imaging models and nonlinear camera imaging models are analyzed and the approximate results of the imaging model and error analysis are researched. A calibration method of camera internal parameters and external parameters based on the planar board is proposed and the results of camera parameters with optical distortions are solved. Measuring device is designed in accordance with the principles of optical-electric, which could realize depth detection of coal level in coal silo under the environment of low illustrate.