Image Coding Using Selective Entropy Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image compression techniques, such as JPEG2000, are computationally intensive due to the use of arithmetic coders for entropy coding, which slows down decompression and compression processes, especially in video applications, while JPEG compression lacks the efficiency of wavelet-based systems and introduces artifacts at low quality.
Innovation Solution
Separate low-entropy and high-entropy portions of quantized coefficients, applying Huffman coding only to the low-entropy portions and leaving high-entropy portions uncompressed, thereby reducing computational complexity without sacrificing compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If arithmetic coders are used for entropy coding in wavelet-based compression systems, then compression efficiency is improved, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent segments the quantized wavelet coefficients into low-entropy portions (most-significant bitplanes) and high-entropy portions (less-significant bitplanes). Different coding strategies are applied to each segment: arithmetic coding is applied only to the low-entropy portions where it provides benefit, while high-entropy portions are handled differently to reduce computational complexity. This segmentation resolves the contradiction by applying the complex arithmetic coding only where necessary.
Solution Approach 2:
The patent applies different coding qualities to different parts of the coefficient data. The low-entropy portions (most-significant bitplanes) receive sophisticated arithmetic coding with detailed bit modeling, while the high-entropy portions use simpler coding methods. This local differentiation optimizes overall compression efficiency while reducing the computational burden of applying complex arithmetic coding to all data.
2Loss of information
If arithmetic coders with sophisticated bit-modeling techniques are used, then compression ratio is improved, but decompression speed deteriorates
Solution Approach 1:
The patent divides the decoding process into two segments: decoding of low-entropy portions using arithmetic coding with bit modeling, and decoding of high-entropy portions using simpler methods. By segmenting the data and applying appropriate decoding complexity to each segment, the patent achieves good compression ratios for the most important data while maintaining faster decompression speeds overall.
Solution Approach 2:
The patent applies partial arithmetic coding - only to the portions of the data where it provides the most benefit (low-entropy portions). This partial application of the complex coding technique achieves sufficient compression efficiency without the full computational overhead of applying it to all data, thereby improving decompression speed.
3Device complexity
If JPEG compression standard is used, then computational complexity is reduced, but compression efficiency and image quality deteriorate
Solution Approach 1:
The patent extracts the entropy-coding step from the complete JPEG pipeline and replaces it with wavelet transform followed by selective arithmetic coding. This extraction allows the patent to adopt the efficient wavelet transform for image decomposition while applying a more sophisticated entropy coding scheme than traditional JPEG, thereby improving compression efficiency without fully committing to the computational complexity of complete wavelet-based systems like JPEG2000.
Solution Approach 2:
The patent changes the parameters of the compression system by using wavelet transform instead of DCT, and by applying arithmetic coding selectively to different portions of the coefficient data. These parameter changes improve compression efficiency and image quality while managing computational complexity through the selective application strategy.
4Loss of information
If wavelet-based compression systems are used, then compression efficiency is improved, but processing time increases due to arithmetic coding overhead
Solution Approach 1:
The patent segments the coefficient data and applies arithmetic coding only to the most significant bitplanes where compression efficiency gains are most valuable. The less significant bitplanes are processed with simpler methods, reducing the overall processing time while maintaining good compression efficiency for the most important image information.
Solution Approach 2:
The patent applies partial arithmetic coding to only the portions of the data where it provides the most benefit. This partial application reduces the processing time overhead associated with full arithmetic coding while maintaining compression efficiency comparable to wavelet-based systems, as the most significant bitplanes contain the most important image information.
Data Source
AI summary
A system that compresses a digital image. During operation, the system performs a transform operation on a digital image to produce transform domain coefficients, and then quantizes the transform domain coefficients to produce quantized coefficients. The system then separates the low entropy portions of the quantized coefficients from the high entropy portions of the quantized coefficients, wherein the low entropy portions of the quantized coefficients comprise the most-significant bitplanes of the quantized coefficients. Next, the system compresses the low entropy portions of the quantized coefficients while not substantially compressing the high entropy portions of the quantized coefficients.