Steel, Cement and Other Industries Use Case

 

Platform Hardware

• Multi-sensory Image Data Fusion (MSDIF) hardware.
• Multi-spectrum Image Fusion (MSIF): Day + VNIR + LWIR

Composite Vehicle Chassis & Sub-systems

CASE STUDY :Object Detection & Tracking

Customer is a Systems Integration Company for delivering weapons to the Indian security force. Customer is focusing on acquiring, in-house design and development of mini and micro UAV systems, complemented by indigenous design of major subsystems like Autopilot, Ground Control Station, and Gimbals based payloads.

• Executive Summary
  Detect and Track (DT) a moving or a stationary object, in an unconstrained live video feed obtained from a remotely piloted unmanned aerial vehicle (UAV), flying over an area of known (Global Positioning System (GPS) enabled) coordinates
• Scope & Challenges
  An object in the field of view (FOV) of the airborne camera, should be automatically detected, filtered and tracked (DFT) in a real time video stream, as per the industry standard of 30 frames per second (fps). The DT algorithm should be invariant to the gradual changes in illumination (day, night and twilight), pose, orientation, speed, scale, occlusion and deformation of the object, image noise and background clutter to an appreciable degree.
• How Solution Helped
  Customer is able to showcase the target identification, tracking and recognition capabilities of its UAV platform to Indian security forces.
  The algorithm multi object, person or event detection and tracking in hyper- spectrum have helped the customer in border surveillance application demonstrations.
• ROI & Future Plan
  Customer is negotiating for a large order of 200 MUAV from Indian Army and has signed a long term agreement of 5 years for supply for Customer wants to have functions of visual navigation odometery and 3D ego-motion. It should measure the relative change in the DT object on the ground with respect to the 3D motion of the UAV.The spatio-temporal coordinates of the target tracked are to be populated in ground control system (GCS) software.
  Challenges
  Integration and optimization of legacy GCS with tracker was challenging as current multimedia framework on which customer GCS is built is not supported after 2009. This affects integrated performance of tracker. We needed to study the legacy system and help customer in integration while optimizing their code base.