A €4 million Horizon 2020 project has begun that aims to improve the performance of embedded image processing applications.
TULIPP, which stands for 'Towards Ubiquitous Low-power Image Processing Platforms', is the European Union’s largest research and innovation programme to date. It is being coordinated by French firm Thales.
By the end of the project in 2018, the hope is that peak performance per watt of image processing applications would have increased fourfold.
The low cost of embedded hardware is set to change the face of machine vision, according to comments made by Jeff Bier, founder of the Embedded Vision Alliance, in an article for Imaging and Machine Vision Europe.
Many machine vision companies are offering solutions that cater specifically for embedded imaging, such as the Imaginghub portal launched by Basler in collaboration with Nvidia, Xilinx and industrial computer manufacturer Advantech.
The focus of the TULIPP project will be to develop a reference platform for vision-based system designers. The platform will define a set of guidelines for the selection of relevant combinations of computing and communication resources, while minimising energy resources, and reducing development costs and time-to-market.
TULIPP will use these guidelines to develop a scalable low-power board designed to meet typical embedded systems requirements of size, weight and power, a low-power operating system and an energy-aware tool chain.
The guidelines will also tackle the design issue complexities surrounding the next generation of embedded image processing applications that are emerging in a range of industry sectors. From an applications perspective, these complexities relate to the need for guaranteed, high performance computing power coupled with greater power efficiency within the context of embedded design requirements.
As well as extending peak performance per watt by four times, the average performance per watt is also expected to increase by 10 times. Beyond the official completion of TULIPP, it is expected that this will be extended to 100 times and 200 times by 2023.
Tulipp will develop three use cases as a proof-of-concept and to validate the reference platform. These use cases will cover different industrial domains with emerging complex image processing requirements and will include: a medical imaging surgical X-ray system designed to significantly reduce radiation doses by 75 per cent; a smart automotive embedded vision system for advanced driver assistance (ADAS) that, in addition to the low-level image processing, interprets intelligently what is on the images to deliver safer driving experiences; and an embedded image processing system to create smart drones and unmanned aerial vehicles (UAVs) for the intelligent search and rescue of survivors at disaster incidents.
‘Image processing applications stretch across an increasingly broad range of industrial domains and are reaching a higher level of complexity than ever before,’ said Philippe Millet of Thales, Tulipp’s project co-ordinator. ‘The TULIPP reference platform will give rise to significant advances in system integration, processing innovation and idle power management to cope with the challenges this presents in increasingly complex vision-based systems.’
The TULIPP consortium members are drawn from both industry and academia and, along with Thales, include Efficient Innovation SAS (France), Fraunhofer IOSB (Germany), Hipperos (Belgium), Norges Teknisk-Naturvitenskapelige Universitet (Norway), Ruhr-Universität Bochum (Germany), Sundance Multiprocessor Technology (United Kingdom), and Synective Labs (Sweden).