Pixel Parameters Optimization in PWM Image Sensor for Quantization Error Suppression

This article established the error model based on pixel parameters to predict the light intensity quantization error in pulse width modulation (PWM) image sensor, attributed to non-linear response, comparator delay and synchronous reading mechanism. The PWM image sensor encodes light intensity into time interval of consecutive pulses, makes the quantization process of light intensity exceptionally impact the accuracy of the imaging quality. Especially in large pixel array, the synchronous readout scheme introduces quantization error. A PWM image sensor with 400(H) × 250(V) pixel array was tested to evaluate the model, and the measurements show that the pixel parameters affect the fluctuation of quantization error. With 300lux illuminance and 25μs reading period, the smallest error fluctuation range (11%-15%) can be obtained. The proposed model can suppress the fluctuation of the quantization error in PWM image sensor and help to achieve improved performance of image quality.