Contact us

A new window will open A new window will open

Scalability beats 'one-size-fits-all' in critical automotive vision systems

Scalability beats 'one-size-fits-all' in critical automotive vision systems

We all know that in many aspects of modern life 'one-size-fits-all' rarely works well. At best it is a compromise, often it adds unnecessary cost and, sometimes, produces a completely unworkable solution.

As manufacturers seek to increase their competitive edge more and more optimised solutions are reaching the market. If you're buying an end product for a specific task, this means you will have something that does the job well - when you're specifying components in a design the optimised solution allows you to transfer the focused performance to your end customers.

Working within the performance-critical and highly competitive automotive sector, Toshiba has created a dedicated, scalable image-recognition processor architecture that makes extensive use of hardware and firmware integration to streamline the execution of frequently-used algorithms, leveraging scalability to handle increasing numbers of applications simultaneously.

The scalable image-processing architecture is built around multiple Media-Processing Engines (MPEs). Alongside these MPEs, hardware filters, transforms and more blocks help accelerate frequently-used algorithms. The first generation of processors to use this architecture featured dual MPEs and a single camera input. The scalability of this architecture has enabled Toshiba to rapidly roll out a family of processors that now provides as many as four MPEs in the second generation of processors and eight MPEs in the fourth-generation devices. With multiple accelerators including affine transform, filters and histogram, as well as multiple camera inputs, the latest devices yield significant increases in simultaneous-processing capability and execution speed.

This scalability enables designers to develop systems that are optimised for the task-in-hand. Lower-end vehicles have less sensors and fewer applications, so a smaller system is implemented. When higher-end vehicles demand more processing power, the designer uses the scalable building blocks they are already familiar with to architect the necessary system quickly and with a high degree of confidence.

To learn how Toshiba's dedicated, scalable architecture will help you meet the most rigorous all-around vision-oriented driver-assistance protocols put forward by Euro NCAP, please click here:

Click here to learn more about implementing ADAS image recognition systems

To Top
·Before creating and producing designs and using, customers must also refer to and comply with the latest versions of all relevant TOSHIBA information and the instructions for the application that Product will be used with or for.