Font Size: a A A

Image Processing Of Underground Coal Mine Based On Compressive Sensing Reconstruction

Posted on:2017-01-15Degree:MasterType:Thesis
Country:ChinaCandidate:M H ZhangFull Text:PDF
GTID:2308330485992874Subject:Control Engineering
Abstract/Summary:PDF Full Text Request
There are a large number of staff in comprehensive mining face field of coal mine. The image information collected under the coal mine is presented to the people in a quick, high quality, high matching degree of reconstruction effect, That has an important guiding significance to coal mining and the safety. In reconstructed image, the traditional method is wavelet transform reconstruction, fourier transform reconstruction, traditional image reconstruction exists many defects, this paper reconstruct image of fully mechanized coal mining face with compressed sensing reconstruction algorithmThis paper is based on the compressed sensing reconstruction of coal mine image processing. In this paper, three core theories of compressed sensing are studied:Signal sparse transformation; Observation matrix designing; Signal reconstruction algorithm. The main studying is three types of reconstruction algorithms of compressed sensing: Convex optimization algorithm; Bayesian statistical optimization algorithm; Iterative greedy algorithm. This paper mainly studies the principle, main steps and flow chart of the three kinds of reconstruction algorithms.In this paper, MATLAB software simulation is used; wavelet transform is used to express the sparse representation of the signal for collected images; Gauss random matrix is used as the observation matrix; five kinds of reconstruction algorithms of compressed sensing is used to reconstruct the original image. These five algorithms are orthogonal matching pursuit algorithm, regularized orthogonal matching pursuit algorithm, gradient pursuit algorithm, base pursuit algorithm and subspace pursuit algorithm. Reconstruction of the same picture with the five reconstruction algorithms, the peak signal to noise ratio of the reconstructed image, the running time and the matching degree of the reconstruction are simulated. The characteristic of the five algorithms for image reconstruction in fully mechanized coal mining face are analyzed. In the application of the actual coal mine image, the appropriate reconstruction algorithm is of great significance to the mining and safety production of coal mine.
Keywords/Search Tags:Compressive sensing, Reconstruction algorithm, Measurement matrix, Sparse signal, Image processing of fully mechanized coal mining face
PDF Full Text Request
Related items