LAUSR

The present study focuses on the performance upgradation of a table-top computed tomography (CT) scanning instrument. The size restriction of the detectors of table-top CT scanner imposes limitation on the allowable size of the object to be scanned. The object with increasing size has to be put in a far end from the X-ray source to be fit into the scan field of view of the scanner. This situation affects the pixel resolution of the tomography images. The limited resolution capacity restricts the CT image zoomed-in view facility which is very often an essential requirement for scientific measurement and analysis. The hardware-based solution involving larger detectors will significantly increase the monetary cost, and thus a cost-effective solution is very much required. The new super resolution (SR) or resolution up-scaling technique based on higher order total variation minimization has been proposed to overcome the above-mentioned problem associated with a table-top tomography instrument. The proposed technique is capable of producing SR image from a single low-resolution image. The experimental results have been compared with some of the existing SR techniques, in terms of various image quality parameters (mean squared error, correlation coefficient, structural similarity index, and feature similarity index) to show betterment over other existing SR techniques. The experiments are carried out with different objects to validate the universality of the proposed scheme. The pixel-enhanced tomographic images are further checked with Sobolev space error analysis to ensure the quality of the images from tomography perspective. It can be inferred that the presented technique can properly up-scale the resolution of CT images, and therefore it presents a cost-effective possible solution to the resolution problem caused by the size constraint of the detectors of table-top CT scanning instruments.