Adaptive exposure control for mitigating rolling shutter effects in CMOS image sensors

CMOS image sensors are prevalent in modern digital cameras due to their efficiency and cost-effectiveness, but they suffer from rolling shutter effects caused by line-by-line image readout. These effects manifest as geometric distortions, especially in scenes with fast motion, due to misaligned exposure moments across sensor rows. This paper introduces an adaptive exposure control method that aligns the mid-exposure moments of all rows through dynamic, row-level exposure adjustment. By varying the start and end capture times for each row, the proposed method ensures uniform mid-exposure timing, thus preserving geometric accuracy in dynamic scenes. Real-time motion analysis is incorporated to allocate shorter exposure times to rows with significant motion to reduce motion blur, while allowing longer exposures in stationary rows to enhance image quality and signal-to-noise ratio. Post-processing compensates for exposure duration variations, ensuring consistent brightness across the image. The approach offers a low-cost and computationally efficient alternative to global shutter sensors, is compatible with existing CMOS architectures, and requires minimal hardware modification. Simulation results demonstrate that this technique effectively mitigates rolling shutter artifacts while maintaining high image quality, making it suitable for consumer devices and other applications requiring robust motion capture.