Home » Technology » Capability

Data Reduction

Deep Vision's data abstraction technology delivers fast and efficient data reduction.

Deep Vision's novel concept of data abstraction quickly transforms abundant sensor data into a form that is easily classified and efficiently stored.

The abstractions created from the raw data are extremely compact. For instance, 64 abstractions (avg. 16 symbols each) requires 1 KB of storage capacity.

By abstracting the raw data into this compact form, the effective storage capacity is increased. For example, a lossless compressed image requires 177 KBα of storage capacity while, the abstractions for the same image would require a mere 1.1 KBβ thus freeing 175.9 KB of storage capacity per image stored.

α Typical storage requirements of a 640 x 480 8 bpp PNG image file.
β Typically. Based on test results using 640 x 480 images.
Exploitation Value

  • Communication between distributed, multi-modal sensors:
    • Counter-mobility Systems
    • UAVs
  • Semantic information storage:
    • Content-based Data Searching
    • Portable Devices

Features

  • Transforms high volume/low value sensor data into low volume/high value information.
  • Efficient information analysis
  • Effectively increases storage capacity
 Operating Facts

  • Operating System: Any (GNU/Linux recommended)
  • Hardware Requirements: None
  • Sensor Modalities: Visual, Thermal, Sonar
  • Timings†: 100+ FPS
  • Runtime Memory Requirements†: 300 KB
  • Storage Requirements‡: 1.1 KB

† Typical. Based on a 640 x 480 data set
‡ Typical. Based on 45 abstractions (avg. 25 symbols each)

Input Requirements

  • Archived data sets
  • Real-time acquisition from visual, thermal, or sonar sensors.