Sunday, March 19, 2017

Analog Devices Unveils Smart Logarithmic Camera

Analog Devices quietly unveils the ADIS1700x, "an embedded vision sensing module engineered for a full range of emerging real-time, smart city and smart building applications, such as parking, traffic monitoring, and municipal light control. The ADIS1700x brings both advanced imaging and intelligence to the edge. The log imager enables clear, accurate imaging in both extreme brightness and darkness. It also generates output contrast at the pixel level, greatly reducing processing load. This allows for on-board algorithms that can deliver at-the-node analytics, so it’s possible to accurately and reliably distinguish between different objects, such as people and vehicles."

ADIS17001 preliminary product page presents a block diagram and the list of features of the new smart camera:
  • Imager: ADVIS200x (QVGA ultrahigh DR logarithmic sensor)
    130dB DR
    Edge detection and homography analytics algorithm for tracking and counting object motion
  • 400MHz ADSP-BF707WCCPZ-4 Blackfin digital signal processing
    Low power of 110mW
    512Mb LPDDR SRAM for advanced algorithms
  • Image Stabilization
    Dual axial digital gyroscope, up to ± 100°/s dynamic range
    Tri-axial digital accelerometer, up to ±16 g dynamic range
  • Autonomous operation and data collection
    Configurable boot-loader installed within flash memory
  • Lens: ADIS17001 provides a 110° horizontal FOV lens, while the ADIS17002 provides a 67° FOV
  • Form factor 45 mm × 45 mm
  • Factory calibrated optical focus and alignment
  • Conformal coated for environmental mitigation

Apparently, the new camera is a result of acquisition of Swiss startup Snap Sensor a year ago.


  1. But the demo image is clearly not QVGA...

  2. This scenario shown here doesn't really need a true HDR camera to combine lights and shadows into one exposure! A true 12Bit camera should be sufficient, with the proper image scaling applied...
    - dkf


All comments are moderated to avoid spam and personal attacks.