Successful video analytics starts with the camera. The right sensor, the right setup, the right lighting are the prerequisites for an image in which you can actually see something. We can advise you extensively on this topic.

For the subsequent image processing and evaluation with AI, we use a framework that allows both on-edge and on-cloud processing. We know the respective architectures, the advantages and disadvantages and the costs.

In doing so, we focus primarily on:

  • Deep Neural Networks (DNN) on Tensor Cores or Tensor Processing Units (TPU)
  • Asynchronous messaging
  • Extensive training data sets

Based on our experience, we either use problem-analogous ready-made DNNs or develop them ourselves.

At the end there is still the presentation or further processing of the information. We provide you with the appropriate user interface, the appropriate data interfaces and also the complete application.