Object Detection

Pleora’s Object Detection plug-in uses sensor data and machine-learning capabilities to identify and classify tanks on a battlefield.

Automatic object detection capabilities can identify and highlight objects that could easily be overlooked by a crew commander focusing on other responsibilities or dealing with limited visibility. The plug-in, which can be trained to identify the type and nationality of a vehicle, automatically processes the image, isolates and visually highlights the object on the commander’s display, and attaches metadata (object name, GPS location, date, time, mission ID).

The commander can press a region of interest on the touchscreen monitor to isolate a single image from the video. Video continues to display on the monitor with the single image overlaid on the left. The image can be quickly categorized by the commander, who can transmit the image to preconfigured recipients (for example other crew members or vehicles) or save the image. The commander can the use video, images, and metadata to warn other vehicles of danger or impassable conditions. Image and metadata can be transmitted to HQ for assistance/back-up and saved for evidence, training, mission reporting.

With Pleora’s plug-in approach to AI, users can train systems on their own unique data and requirements. AI is particularly well-suited to locate, identify, and classify objects in moving scenes to decrease sense to response time for drivers and crew commanders. Pleora’s drag-and-drop AI development platform provides a straightforward approach to developing custom algorithms, with off-the-shelf sample frameworks that can be trained using your unique data to speed development and deployment. Plug-ins are optimized for Pleora’s RuggedCONNECT Smart Video Switcher and flexible for other processing choices.