Helmet detection for large industrial sites
Kapernikov developed a Proof of Concept (PoC) for a worldwide energy company, based on deep learning ...
Machine vision
Automate any visual task
Kapernikov uses sensors to allow your machines to detect objects in an image, estimate the position of these objects or handle them with care. All of these human capabilities have become available for machines through artificial intelligence.
Computer vision relies on a wide range of sensors, from hyperspectral cameras to time-of-flight sensors. By turning sensor data into useful information, we help you make better automated decisions.
You can use imaging sensors for a wide range of robot applications or process monitoring setups:
We configure the capture set-up, select the lens and the sensor and we develop the algorithms. No matter how challenging the application, we help you decide which approach to use: either a classic approach or deep learning, training in the cloud and deployed on an edge device.
We build camera-based positioning solutions to locate people or objects in manufacturing environments, in warehouses or on building sites. We use existing security camera networks or deploy alternative optical measurement systems like LiDAR or laser for harsh environments. Other technologies like Ultra-Wide Band (UWB) positioning can also optimize your roll-out cost.
We use state-of-the-art sensor technology and data analytics to allow Automated Guided Vehicles to move efficiently and safely on the production floor. We combine visual odometry, radio positioning (UWB) and LiDAR with robot signals like speed and steering angle.
We typically deliver turn-key projects for manufacturers. But just as much, we like to share our computer vision know-how with our customers. For machine builders, our consultants can be the ideal in-project support. We share our knowledge about the latest evolutions in hardware and software and work together with your team to shorten your development cycle.