Detection of coal dust in a mine using optical tomography

A dust concentration imaging system based on optical tomography is proposed to monitor concentration variations of coal dust in a mine. Concentration profiles, rather than just a point value, of coal dust concentrations are the goal of this method. An optical sensor array is employed to realize an optical sensing field of the coal dust concentration during on-line monitoring. A novel image reconstruction algorithm, called the simultaneous iterative reconstruction technique (SIRT), is compared to the well known linear, back projection algorithm (LBP). The SIRT was applied to the present problem and tested by modeling. The disadvantage of the SIRT is a slow speed but some improvements have been made by adding a weighting function that reduces the relative error to 1.7% from 3.1% for 50 iterations. The results of image reconstruction are presented for both simulated and real objects. They prove that the optical tomography technique, based on a multi-source fan projection scheme, can be an effective approach for estimating coal dust distribution. This system can be applied in real time for continuous measurements in a mine.