New architectures for multi-camera operations in vehicles

1st March 2019

Alex Lynn

0 0

Today’s surround view systems may have four cameras or more to observe and analyse the scene around an automobile. Typically, at least four camera streams will be stitched together in a 2D or 3D display for the best driver experience. Camera monitor systems (CMS), may require two cameras on the driver’s side and two cameras on the passenger side, rendering the video streams to create a very wide view.

Both architectures interconnect megapixel image sensors with one central application processor by bridging the long distance over a single coax cable. This requires that the video streams to arrive fully synchronised. While current generation systems are one megapixel (MP) in resolution, but next-generation systems demand image resolutions of two megapixel and more at up to 60 frames per second frame (fps) rate.

Light-emitting diode (LED) flickering caused by the pulse-width modulation (PWM) of LED lights in traffic signs and cars is another problem. In addition, the market demands higher dynamic range (HDR) imaging suitable for extreme lighting conditions, requiring additional pixel processing.

Video stream transmission systems need to feature central clock distribution; video synchronisation, back-channel communication and power over coax (PoC) capabilities. In order to achieve all this, highly integrated solutions feature chipsets of multiple FPD-Link serialises combined with a single FPD-Link deserialiser hub.

This paper from Texas Instruments discusses how to fulfill market demands for HDR, LED flicker mitigation, low-light performance and advanced image quality.

To read the full paper, click here.