Image Sensors for Camera Monitor Systems

Abstract

Image sensors are a critical part of camera monitor systems (CMS). Photons passing through the imaging lens impinge upon microlenses that focus the light through color filters and onto the sensor pixel photodiodes. Two principal sensor architectures in use today are front-side and back-side illumination (FSI and BSI), the latter of which is just becoming available in the automotive space. Pixel signal integration and readout can be implemented by electronic rolling shutter or global shutter mechanisms; the former dominates viewing applications but the latter can be advantageous for machine vision. The signals are read out by analog circuitry, and converted to the digital domain. The digital signal is processed to produce an image suitable for display or analysis. This process includes steps that: correct for black level and white balance; fill in missing color signals via a demosaic algorithm; spatially process the image to distinguish signal and noise; map the signals to a standard color space; and apply tone-scaling. Critical sensor performance measures for CMS applications include signal to noise ratio (SNR) and modulation transfer function (MTF); the advanced Fourier metric noise equivalent quanta combines these two measures and is useful in understanding task performance. Scene dynamic range in automotive applications can be very high, beyond the capabilities of a single pixel exposure, and therefore special techniques are required. The mitigation of flicker from time-varying light sources, such as fluorescent lamps and light emitting diodes (LEDs), is a current challenge, particularly in high dynamic range (HDR) imaging. To exemplify the practical implementation of image sensors in CMS, a sample system design is presented. Image sensors with integrated safety features support CMS implementations that require ISO 26262 automotive safety integrity level (ASIL) ratings. CMS is an exciting new application, with considerable opportunity for innovation, particularly in the image sensor and processing components. Advances in these areas will contribute to the greater adoption of CMS systems in vehicles in coming years.