CoSi

Summary

Evaluate colour differences is of great importance for many applications, particularly when image processing or computer vision techniques are employed.

The CoSi technology can evaluate colour similarity, i.e., comparing two colours using human perception criteria (CIE 2000), reducing up to 95% the required processing time with the same accuracy.

Challenge | Opportunity

The usage of colour similarity algorithms has been increasingly important in several fields such as Multimedia, Healthcare, Agro-food, or Materials, which require image and video analysis and processing, content management, and real colour comparison. Colour identification presents challenges mostly related to the image’s noise and illumination and the computational cost, complexity, or time to comply with recommended metrics (CIE 2000). These limitations hinder real-time or portable applications. 

INESC TEC developed an improved method to calculate CIE 2000 needing just 5% of the computational power usually required by the conventional implementation of this standard. This improvement allows determining CIE 2000 in real-time video as well as in portable contexts.

Main Advantages

• Speed
• Low processing complexity
• High accuracy

Applications

• Image segmentation (as a critical step in image and video processing algorithms)
• Digital image processing
• Computer vision
• Specific industries:
  • Entertainment
  • E-commerce
  • Health
  • Agro-food
  • Materials