A fast fractal image coding based on kick-out and zero contrast conditions

Area of Science:

• Computer Science
• Image Processing
• Data Compression

Background:

• Fractal image coding offers high compression ratios but often suffers from high computational complexity.
• Existing fast fractal algorithms may require pre-processing or additional memory, limiting their practical application.

Purpose of the Study:

• To develop a computationally efficient fractal image coding algorithm.
• To reduce the runtime and memory requirements for fractal image compression.
• To maintain high reconstructed image quality comparable to traditional methods.

Main Methods:

• A novel algorithm employing a single kick-out condition to minimize range-domain block matching.
• An efficient zero contrast prediction method for exact computation of block differences.
• Integration with other fast fractal algorithms, such as the DCT Inner Product algorithm.

Main Results:

• Achieved a 50% reduction in runtime compared to exhaustive search methods.
• Demonstrated equivalent reconstructed image quality to exhaustive search.
• Showcased superior efficiency and reduced memory footprint compared to other fast fractal algorithms.
• Further runtime reduction of approximately 50% when combined with the DCT Inner Product algorithm.

Conclusions:

• The proposed fast fractal image coding algorithm offers a significant speed-up without compromising image quality.
• The algorithm is memory-efficient and does not require pre-processing, making it practical for implementation.
• This approach provides a strong foundation for further advancements in efficient fractal image compression techniques.