Homography Transformation Correction Method for Position Error Generated in Readout Circuit Based on Resistive Network for the Compton Imaging System

A homography method to correct position errors generated in the Compton imaging system using a resistive network is presented. The Compton imaging system is composed of a scatterer and an absorber in multichannel arrays for high resolution and can detect gamma rays emitted from radioisotopes. Resistive networks are often used in this system to efficiently reduce the number of channels. However, this can cause position errors, and the spatial resolution deteriorates according to the resistance value of the network, type of detector array, and characteristics of the preamplifier used. Therefore, before tracking the position of the source, it is necessary to correct the position errors of images obtained from the scatterer and absorber. Also, a new correction method should consider the characteristics of the readout circuits based on the resistive network. In this work, the position errors are corrected using homography, which is a coordinate transformation method. To verify the corrections using homography transformation, we modeled the current pulse generated from the detector and designed an automatic channel selection circuit to input each channel of the resistive network. From experiments, we first obtained the positions with distortions according to the setup of readout circuits and corrected these errors by applying the homography transformation method. Consequently, the distortions were significantly corrected, and the error rates of the positions compared with those of the ideal grid were greatly reduced by up to 0.36%.