MIPI Update Enables Next-Gen Always-On Machine Vision Applications

The MIPI Alliance issued an update to its MIPI Camera Serial Interface 2 (MIPI CSI-2) that enables advanced machine vision applications in areas including mobile, augmented and virtual reality, drones, IoT, medical devices, and automotive.

Building on the machine awareness capabilities introduced in MIPI CSI-2 v3.0, v4.0 adds an advanced always-on imaging solution that operates over as few as two wires to lower cost and complexity for ultralow-power machine vision applications. It can continuously monitor the surrounding environment and wake its host CPU when significant events are detected.

Example uses include laptop or tablet-based face sensing, video surveillance, and vision-based vehicle safety applications. Image frames can be economically streamed from an image sensor to a video signal processor over a low-power MIPI I3C bus, with scaling options to add extra I3C lanes and bandwidth as defined by the I3C specification.

The MIPI Alliance introduced a significant update to its MIPI Camera Serial Interface 2 (MIPI CSI-2) interface to enable advanced machine-vision applications in multiple application spaces, such as mobile, augmented and virtual reality, drones, the Internet of Things (IoT), medical devices, industrial systems, automobiles, and client devices such as tablets, notebooks and all-in-ones. Building on the machine awareness capabilities introduced in MIPI CSI-2 v3.0, v4.0 adds an advanced always-on imaging solution that operates over as few as two wires to lower cost and complexity for ultra-low-power machine vision applications. Courtesy of the MIPI Alliance.

CSI-2 v4.0 also adds multipixel compression for the latest generation of advanced image sensors and RAW28 color depth for better image quality and signal-to-noise ratio. The system provides optimized pixel compression for the latest generation of Tetra- and Nona-Cell image sensors with multipixel color filter arrays (CFAs).

The feature helps alleviate the high bandwidth demands of these sensors by compressing both multipixel and standard Bayer CFA images more efficiently and with potentially higher quality than current methods. Reference code has also been made available to help integrators evaluate the algorithm and verify their implementations.

CSI-2 v4.0 is backward compatible with all previous versions of the MIPI specification.