Computer vision

 Computer-vision

Helping Businesses See into the Future

Computer vision – the use of algorithms on input from a camera or other sensors to extract usable data for the device.

Our experience in integration and development of Embedded Systems based on Vision and sensor fusion algorithms is:

  • Object detection & classification
  • 3D-reconstruction
  • motion estimation & tracking
  • image processing on IPU

Our challenge: squeezing the model to fit on an Embedded board

Optimization for designated compute units / hardware accelerators:

  • Vectorization primitives on CPU and co-processors. For example, ARM NEON.
  • General Processign on Graphics Processing Unit (GPGPU) using CUDA and OpenCL.
  • Image Processing Unit (IPU)
  • Video Signal Processor (VSP)
  • Specialized Digital Signal Processor (DSP) and Field-Programmable Gate Array (FPGA)

We support and integrating industry standard frameworks such as:

  • libLAS
  • OpenCV
  • OpenSLAM
  • ROS

Our expert application group can assist clients with integrating Computer vision algorithms for any purpose, implementing these algorithms, and assuring optimized performance with hardware.

 

L4B
Where Creative Ideas and Technical Expertise Produce Reliable Business Products