Deep Analytics | Case Study

Video Intercepting Depth Overlay (VIDO)


The Video Intercepting Depth Overlay, or VIDO, is a unique prototype designed to inject sensor data into existing video streams on ground robots. Requirements dictated that the system inject four streams with sensor data, and that video streams could be analog, digital via USB, or IP-based RTSP.

Because the customer was particularly cost sensitive, we developed a custom motherboard capable of interfacing with 4 ultra-low-cost (~$20) card edge computing devices. Each computing device natively possessed USB, ethernet, and analog outputs. Custom electronics were developed to capture frames from input analog signals, and a custom Linux distribution was developed to autonomously emulate attached USB cameras and RTSP devices. Sensor data was written onto a digitized video stream and provided output back to the host system with virtually no perceptible delay in the video stream. The custom motherboard handled power, high speed signals, card edge connections, signal conditioning, and connectorization[ak1]  for sensor and video inputs and outputs.