Abstract
It is an important step to generate position profiles for block detectors in the calibration process of a Positron Emission Tomography (PET) scanner. Automatic peak detection is desirable because manually selecting each peak is time-consuming, especially for large crystal arrays (e.g. 20x20). In this project, an edge-based approach is proposed. The proposed algorithm utilizes the information on the edges and identifies the region that containing a single light spot. Then, the peaks are detected in each region. A GUI tool is programmed for users to visualize the result and allow the users to correct each peak manually when there is any error during the automatic peak search. The algorithm and tool have been tested by 61 sample images. The result is satisfactory with average accuracy over 95%.