KOLO EDGE
AI module
Coming soon
KOLO EDGE AI module turns any camera system into a high-performance autonomous sensing module. It allows processed AI detection, identification, and tracking capability on any camera system.
KOLO EDGE AI module can be used for all types of new cameras and upgrade existing ones.
Integration is done by plugging the KOLO EDGE between the camera and communication equipment.
The model training unit will automatically adopt the Neural Network to the current installation place and update it based on the data received by the system operation.
General capabilities
Autonomous capabilities using Neural Networks:
-
Detection of land, air, and sea targets
-
Classification of detected objects
-
Tracking of fast-moving objects
-
Up to 360º panoramic view
-
12 predefined classes in the classification model as well as custom user classes
-
Additional image stabilization and enhancement algorithms
System additional capabilities:
-
Plug-and-play integration to any IP-based camera system
-
Training during the system operation for adopting a system to the current installation place enhancing the neural network model
-
Processing IR camera as well as Day/Night camera
-
Lower requirements to data link due to image processing on the edge
-
Rudgered case for outdoor installation
Extended capabilities with edge AI processing
Object detection/identification:
-
Neural network object detection/identification
-
Autonomous false positives rejection
-
12 predefined classes for air, land, and sea targets
-
Edge processing allows to use of the image in maximum quality
-
Edge processing enables object detection capabilities without replacing existing user infrastructure
Object tracking:
-
Deployment tracker on the edge device allows for tracking of fast-moving objects.
-
Object tracking based on AI provides better tracking accuracy compared to the common trackers.
-
AI tracking based on Siamese family on neural networks.
Real-time panoramic scanning:
Real-time panoramic scanning allows coverage of large sectors with one camera. The camera moves in the predefined sector from side to side simultaneously processing each frame for object detection. The operator will have the possibility to observe all areas at one time with all options for AI object detection.
That allows to provide 360º situation awareness, on a single 2D operator’s monitor.
No-code system programming
No-code software employs a visual development interface, allowing individuals without technical expertise to design complete AI applications through the intuitive process of dragging and dropping software elements. With this tool, users can fashion customized workflows aligned with internal needs and even train new neural network models.