With multiple sensors per system and growth in sensor diversity coupled with sensor processing demands and many proprietary APIs, the cost and time needed to integrate and utilize sensors in embedded systems is also becoming more challenging.
To address this, The Khronos Group, an open consortium creating graphics and compute interoperability standards, and the European Machine Vision Association (EMVA), a European industry association dedicated to vision technology, have announced the formation of an embedded camera API exploratory group, open to all at no cost, to explore industry interest in the creation of open royalty-free API standards for controlling embedded cameras and sensors.
All participants will be able to discuss use cases and requirements for new interoperability standards to accelerate market growth and reduce development costs in embedded markets using vision and sensor processing and associated acceleration. If the exploratory group reaches significant consensus then Khronos and EMVA will work to initiate the proposed standardization projects at the appropriate organizations.
All sensor and camera manufacturers, silicon vendors, and software developers working on vision and sensor processing are invited to participate in this initiative. More details and instructions for joining the group are here.
The embedded camera API exploratory group has been created in response to industry requests. Increasingly, camera sensors are being tightly integrated with image, vision and inferencing accelerators in self-contained systems. Innovation and efficiency in the embedded vision market is becoming constrained by the lack of open cross-vendor camera control API standards to reduce development and integration costs of multiple advanced sensors and cameras.
A consistent set of interoperability standards and guidelines for embedded cameras and sensors could streamline deployment by manufacturers and system integrators by enabling control of a wide range of camera sensors, depth sensors, camera arrays and ISP hardware to generate sophisticated image streams for downstream processing by diverse accelerators.
This exploratory group will use Khronos’ proven framework for new initiatives in collaboration with the EMVA. Any companies, universities, consortiums, open-source participants, and industry experts who are willing to sign an NDA are welcome to join, at no cost. All participants will have an equal voice in exploring industry needs for, and benefits of, creating a consensus to develop a scope of work (SOW) document describing the objectives and high-level direction of standardization initiatives of value to the industry. The group is expected to meet online over a period of several months starting on March 25, 2021.
All exploratory group discussions will be covered by a simple project NDA to encourage open discussions. The group is open to all proposals and relevant topics but will not discuss detailed technical design contributions to protect participants intellectual property (IP). If a SOW is agreed, Khronos and EMVA will work to initiate the standardization work at the most suitable host organizations or open source projects, using those organizations’ normal collaborative agreements and IP frameworks.
Those who have already indicated an interest in joining the exploratory group include ALL3D, Almalence, AMD, Apertus, AREA, Arm, Cadence, Codeplay, Collabora, EA, Facebook, Google, Holochip, HP, Huawei, LunarG, Mobica, Nvidia, Oculus, OPPO, Qualcomm, RedHat, Texas Instruments, Ultraleap, and Valve from Khronos; as well as EMVA members and machine vision players such as Allied Vision, Basler AG, Baumer, MVTec, and Stemmer Imaging AG.
“Judging by the significant industry interest, the time seems right to organize an effort around identifying and aligning on the need for interoperability APIs for embedded cameras and sensors. This is a topic that is very relevant to Khronos as our acceleration APIs, such as OpenCL, SYCL and OpenVX are often used to accelerate sophisticated sensor stream processing,” said Neil Trevett, Khronos Group president. “Our work is also very complementary to EMVA, and we are delighted that the two organizations are working together to bring a meaningful quorum from diverse parts of the industry into this cooperative exploratory process.”
“We are delighted to work with Khronos on this initiative to commonly understand the industry needs for the future of embedded vision,” said Dr. Chris Yates, EMVA president. “Both the EMVA and the Khronos group have a well-established history of standardization developments which enable industry to develop new products more simply, whilst ensuring friction is reduced in the market. This Exploratory Group is an excellent approach to understanding broader industry needs and will bring together many companies and views in an open forum.”
Industry response has been supportive, according to the two organizations.
“Embedded vision is a natural progression from full-sized PC-based vision systems to systems on a chip and is critically important to the future of the vision industry. The industry has seen great benefits from digital interface/interoperability standards such as GigE Vison and USB3 Vision in expanding markets, reducing costs, and simplifying technology application. It makes great sense to continue these standardization concepts at the embedded level,” said Jeff Burnstein, president of the Association for Advancing Automation (A3), parent association to AIA – Advancing Vision+Imaging.
Christine Perey, interoperability and standards program leader for the Augmented Reality for Enterprise Alliance (AREA), said, “Lack of API standards for advanced use of embedded cameras and sensors is an impediment to industry growth, collaboration and innovation. Enterprise AR customers and systems integrators/value added providers will benefit from greater clarity, open interfaces between modular systems and innovation in the component provider ecosystem. Standards for camera and sensor control will increase opportunities for powerful new combinations of sensor and AR compute resources, integration with existing IT, and lower cost and complexity of future solutions.”
- New ams image sensors for high throughput industrial vision
- AI vision processor enables 8K video at 30fps in under 2W
- Arm adds CPU, GPU and ISP for autonomous and vision safety
- Smart camera offers turnkey edge machine vision edge AI
- Embedded Vision Summit: ML perception on tiny FPGAs
- CoaXPress 2.0 chips enable machine vision at 12.5Gbps on single cable
- FPGAs displace ASICs in Subaru Eyesight vision-based ADAS