Machine vision cameras can be set up in production lines or mounted on moving parts such as robot arms in order to acquire images in industrial settings for such purposes as checking manufacturing tolerances, localization and geometrical measurement of objects, remote monitoring, etc. Such cameras can be integrated into an environment with external lighting or provide their own light source, and may cover different spectral bandwidths (RGB, IR, UV, etc) and sensitivities to light.

They are typically expected to be powered on and available at all times for extended timespans measured in years, sometimes in difficult conditions (extreme heat or cold, vibrations, etc). Controlling the image quality of the camera system is critical to guaranteeing accuracy and repeatability of the installation, as well as to quickly identifying failure modes over time and thus avoiding productivity losses. Analyzer’s flexible testing charts come in different sizes, from close-range inspection charts that can be integrated inside an inspection line to large-scale charts mounted on walls or on easy-to-move wheeled easels.

The positioning markers on such charts facilitate obtaining automated measurements of sharpness, noise, color rendering, geometrical distortion, and frame rate (among many other metrics) in any condition. The Analyzer measurement software suite provides plenty of options for evaluating precise image quality metrics at any resolution, and in both still image and video formats.

By integrating our Python API into your systems or even by building custom testing solutions with our unitary measurement tools, you can test your existing cameras with easy-to-use measurement reports and documented numerical guidelines; evaluate, check and compare specifications for off-the-shelf camera systems with standards-compliant measurements; and automatically receive alerts whenever image quality parameters deviate from their nominal values.

Use Cases:
  • Calibration and tuning in situ require “small”
    targets like the P2020 SFR chart
  • High precision of pixel positioning requires a highly
    accurate distortion model with error estimation
  • Noise evaluation according to the emva1288
    standard
Measurement and KPIs:
  • Chromatic aberrations
  • MTF
  • Noise
  • Tone curve
  • Color fidelity
  • Precise distortion calibration (Brown-Conrady
    models compatible with OpenCV and Matlab) with
    error map for accuracy evaluation
Starter Kit: