Skip to main content

AI Image Analysis

The AI Image Analysis of Viesus examines an image to detect the conditions and technical properties of the image as well as the individual characteristics of the subject.

The AI Image Analysis informs the subsequent processing steps, such as Image Optimization, Upscaling and Repair, and provides the necessary data for their granular adjustments.

Technical Properties and Image Conditions​

The technical properties and image conditions include the detection of:

  • Color distribution - Analysis of color balance and saturation across the image
  • Tint - Detection of color casts and hue shifts
  • White balance - Evaluation of color temperature accuracy
  • Contrast - Assessment of tonal range and dynamic range
  • Brightness/exposure - Detection of under or overexposed areas
  • Sharpness - Analysis of image clarity and focus quality
  • Noise/grain - Identification of digital noise and film grain
  • Image size/resolution - Evaluation of pixel dimensions and quality
  • Red eyes - Detection of red-eye artifacts in flash photography

Characteristics of the Subject​

The analysis of the individual characteristics of the subject include:

Scene Classification​

  • Indoor/outdoor detection - Determines shooting environment
  • Scene types - Landscape, portrait, street photography, macro, etc.

Image Content Analysis​

  • Environmental elements

    • Sky regions
    • Vegetation and foliage areas

  • Subject detection

    • Number of faces and their relative sizes
    • Face positioning and orientation
    • Skin tone analysis and values

Lighting Conditions​

  • Shadow and highlight analysis - Detection of tonal extremes
  • Back-lit subjects - Identification of challenging lighting scenarios
  • Light direction - Assessment of primary light source positioning
  • Ambient lighting - Analysis of overall scene illumination

Processing Integration​

The AI analysis results directly influence:

  • Enhancement algorithms - Tailored adjustments based on detected conditions
  • Noise reduction - Optimized for specific noise patterns and levels
  • Color correction - Targeted fixes for detected color issues
  • Face enhancement - Specialized processing for portrait images
  • Upscaling decisions - Algorithm selection based on content type