Driver Assistance (DA) systems support drivers and help in reducing accidents and associated fatalities caused by distraction and human errors.

We are working with our partners to provide video processing algorithms suitable to use as core components of FPGA-based DA systems. These IP blocks are the result of years of R&D and address those automotive system integrators needing to reduce the time to market of their products starting from performing, reliable and flexible solutions.

Why FPGA?

There is strong interest from many industrial organizations to improve automotive safety through DA systems development, which is considered a very challenging task from both scientific and technological point of view.

The emerging market for automotive DA systems requires:

  • high-performance digital signal processing
  • low device costs appropriate for a volume application
  • short development time

The Xilinx FPGA devices represent a suitable platform to meet all of these requirements. In particular the Xilinx Zynq®7000 All Programmable SoCs, combining programmable logic and high-performance embedded processor on the same chip, provide the most effective solution for targeting DA algorithms demanding both intensive pixel-level calculus and high level complex control.

eVS is member of the Xilinx Alliance Program.



Detecto Object Detector

This is a vision-based object detection IP core. The algorithm follows a discriminative approach combining a HOG-based descriptor and a SVM classifier fitting, in particular, pedestrian detection. It supports multiple scales to detect objects moving in an arbitrary range of distance. >> More...

DPM Deformable Part Model Detector

DPM is an object detection IP core implementing the deformable part model algorithm in FPGA. It uses a star-structured part-based model, defined by root filters plus a set of parts filters and associated deformation models. It supports multiple scales to detect objects moving in an arbitrary range of distance. >> More...

VDET Vehicle Detector

This IP core is a learning-based vehicle detection, developed for embedded vision applications. The algorithm follows a discriminative approach based on a cascaded classifier using Local Binary Pattern features. It is trained for recognizing rear view of cars and trucks. >> More...

 

Rear-Looking Lane Marking Detection 

This IP core is designed to detect the lane markings on the roadway video scenarios captured from a rear view wide angle camera. It can be used as base component of a rear-looking Lane Departure Warning (LDW) system. With respect to a forward-looking camera system it captures images on a shorter range so a simpler lane model can be adopted.>> More...

 

Forward-Looking Lane Marking Detection and Tracking

This IP core is designed to detect the lane markings on the roadway video scenarios captured from a camera installed in the front of a vehicle. It is the base component of a forward-looking Lane Departure Warning (LDW) system whose objective is to alert the driver when the vehicle inadvertently strays from the road lane. >> More...

Backend Options

See here all the options you can set in the Joomla Backend of this Theme.

Backend Options

×

Login


eVS :: embedded Vision Systems :: Registration



×

Register



  or   Login
×