We evaluate the performance of an Omnivision OV5647 CMOS image sensor (5 Mp) for detecting radiation from Sr90 and Cs137 sources. Our experimental setup uses a Raspberry Pi 3 mini-computer for data acquisition, with image processing using Python and OpenCV libraries. We specify the necessary settings to convert a standard camera into a particle detector sensitive to electrons and photons, including a two-step background filtering procedure. In addition, we implement the first detailed Geant4 simulation that describes the layered geometry and material composition of a commercial CMOS sensor along with the radioactive sources. To enhance the simulation, we include an algorithm for charge diffusion and conversion of the energy deposited by electrons and photons into ADC counts. Our measurements are presented in terms of cluster size, the maximum ADC signal per cluster, and the number of clu...

The authors gratefully acknowledge the Dirección de Gestión de la Investigación (DGI-PUCP) for funding under Grant No. DGI-2019-3-0044. C.S. acknowledges support from the Peruvian National Council for Science, Technology and Technological Innovation scholarship under Grant No. 236-2015-FONDECyT. We would also wish to thank R. Sanchez from the Applied Optics Group and the Quantum Optics Group for letting us use their lasers and equipment, as well as Y. Coello for the photodiode. We also thank J. A. Guerra for useful discussions and suggestions.