Optimization of photon counting algorithm based on ICMOS detectors

ICMOS detectors usually use analog integration method to process signals in the low-light field, but when the light intensity is reduced to the single-photon level for very-low-light detection, the detection sensitivity and Signal-to-Noise Ratio (SNR) are significantly degraded due to the limitations of low photoelectric conversion efficiency and system noise. To enhance the performance of ICMOS detectors in very-low-light detection, this study introduces photon-counting imaging through algorithmic optimization and improves the single-photon event resolution accuracy through noise suppression to optimize the quality of photon-counting imaging. In this paper, a three-step photon counting algorithm optimization scheme is proposed: firstly, the background noise level under different light intensity conditions is optimized by combining the system characteristics with Gaussian fitting and dynamic threshold adjustment of adjustment factor $k$; secondly, the signal cluster response are extracted by combining the connected component analysis and the 8-connected seed-filling algorithm, and the residual background noise, which is manifested as the point response, is removed; and lastly, the single-photon event resolution is performed by the centroid method. In addition, in order to verify the effectiveness of the optimization scheme, this paper compares the optimized photon counting algorithm with the traditional analog integration method, and verifies the improvement of the SNR by photon counting statistics and contrast calculation. The results show that the optimized photon counting algorithm not only significantly improves the detection sensitivity and SNR of the ICMOS detector, but also effectively improves the spatial resolution.