Fujitsu Semiconductor Releases 'Approaching Object Detection Library' to Support Driver Safety
(ENP Newswire Via Acquire Media NewsEdge) ENP Newswire - 16 May 2014
Release date- 15052014 - Yokohama, Japan - Fujitsu Semiconductor Limited announced that it has released the Approaching Object Detection Library, software for use with the company's graphics SoCs to visualize information from vehicle-mounted cameras to detect people, vehicles, and other objects nearby.
This uses proprietary technology from the company to dramatically reduce detection errors, and because it has been optimized to run as software, it can be implemented without dedicated hardware that is typically necessary. An evaluation version of the Approaching Object Detection Library will be available beginning in May, with a commercial version due in September.
Fujitsu Semiconductor has been developing and delivering visualization solutions for vehicles that take advantage of the high speed operation and high quality graphics capability of the company's high-performance SoCs for vehicles, such as integrated Human Machine Interface (HMI) systems that consolidate and provide centralized control over multiple displays, including cluster (dashboard) displays, central console displays, and heads-up displays, and the 360-degree Wraparound View System, which synthesizes 360-degree images from multiple vehicle-mounted cameras to enable drivers to check their entire surroundings in 3D from any angle.
The Kids and Transportation Safety Act in the US will require that all passenger cars made from May 2018 on include rearview cameras. Because of this and other reasons, demand for functionality that can provide drivers with information on their vehicles' surroundings is expected to significantly increase. Approaching-object detection functionality that will automatically detect nearby cars and people, and alert the driver, is an extremely effective method for lightening the driver's cognitive load, and there has been a need for products that easily implement this functionality without sacrificing performance.
Fujitsu Semiconductor regards approaching-object detection functionality as an important part of its visualization solutions for vehicles, and developed the Approaching Object Detection Library with that in mind. The product uses the company's proprietary technology for reducing detection errors to dramatically cut down on false positives and false negatives. In addition, the functionality is implemented to run as software, with parallelized data processing and other optimizations. Use of this product enables high-performance and flexible approaching-object detection systems to be built without dedicated hardware such as image processing LSI devices.
The Approaching Object Detection Library will be available in an evaluation version in May this year, when it will be shipped with an evaluation board that includes the company's graphics SoC, MB86R24, for testing functionality. Release as a commercial product is due in September. Fujitsu Semiconductor aims to take the world's leading share of the market for products in this field that use this kind of technology.
Through its visualization solutions, Fujitsu Semiconductor aims to help build a society where cars co-exist safely with people. The company also plans to apply the technology it has developed for cars onto general SoC platforms, and to develop many new applications, such as for surveillance cameras and automatic controls.
The company will be demonstrating the Approaching Object Detection Library and other solutions at the 2014 Automotive Engineering Exposition, running May 21-23 at the Pacifico Yokohama.
The Approaching Object Detection Library: Operational Summary
The library generates the optical flow (optically detected movements) within the input video stream, applies detection-error reduction processing, and then groups objects with similar flows to identify moving objects.
The Approaching Object Detection Library: Features
Implements approaching-object detection functionality in software
In this library, the analysis of images from vehicle cameras and the detection algorithm that identifies approaching objects are optimized to run as software, while image processing is accelerated through such techniques as parallelized data processing. This makes it possible to dispense with dedicated hardware for image processing, and allows for more flexible configuration to fit into a variety of conditions and applications.
Proprietary detection error reduction technology
Fujitsu Semiconductor developed a new algorithm that limits detection errors such as false positives and false negatives. By identifying the factors that lead to those detection errors and correcting them through the validation of a massive volume of images, this technology greatly improves detection performance with fast-moving objects and in low-light conditions.
CPU ARM Cortex-A9 533MHz
ROM 90 Kbytes
RAM 4 Mbytes (input image size: 720x480, detection area: 2 areas of 240 x160 pixels)
Product Name Availability
Approaching Object Detection Library September 2014
Please contact your sales representatives for sample price.
About Fujitsu Semiconductor
Fujitsu Semiconductor Limited designs and manufactures semiconductors, providing highly reliable, optimal solutions and support to meet the varying needs of its customers. Products and services include Customized SoCs (ASICs), Foundry Service, ASSPs, and Ferroelectric RAMs (FRAMs), with wide-ranging expertise focusing on mobile, imaging, automotive and high performance applications. Fujitsu Semiconductor also drives power efficiency and environmental initiatives. Headquartered in Yokohama, it was established as a subsidiary of Fujitsu Limited on March 21, 2008. Through its global sales and development network, with sites in Japan and throughout Asia, Europe, and the Americas, Fujitsu Semiconductor offers semiconductor solutions to the global marketplace.
For more information, please see: : http://jp.fujitsu.com/fsl/en/
Fujitsu Semiconductor Limited
Fujitsu Semiconductor Limited
(c) 2014 Electronic News Publishing -
[ InfoTech Spotlight's Homepage ]