Block Modulating Video Compression for Mobile Power Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video and image codecs require high computational complexity and power consumption, making them unsuitable for mobile devices with limited resources, which often resort to lossy compression methods resulting in material quality loss.
Innovation Solution
A block modulating video and image compression (BMVC) method that divides an input image into blocks, applies different masks to each block, and sums the modulated blocks to generate a single encoded block, reducing computational complexity and power consumption while maintaining image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If existing high-quality video codecs are used, then image quality is improved, but computational complexity and power consumption increase
Solution Approach 1:
The patent divides the input image into multiple blocks (e.g., 8x8, 16x16 pixels) and applies different random masks to each block independently. This segmentation allows the encoder to process smaller units with simpler computations while maintaining overall image quality through the combined effect of all blocked transformations.
Solution Approach 2:
The patent introduces random mask parameters (binary masks with 0s and 1s or grayscale masks with values between 0 and 1) that are applied to modulate pixel values in each block. These parameter changes transform the image data in a way that enables compression while preserving quality, as the masks create controlled variations that can be reversed at the decoder.
2Manufacturing precision
If existing high-quality video codecs are used, then image quality is improved, but power consumption increases
Solution Approach 1:
By segmenting the image into blocks and applying simple random masks to each, the patent reduces the computational burden on mobile devices. The encoder performs only basic multiplication operations (pixel value × mask value) rather than complex transform and quantization operations, significantly lowering power consumption while maintaining acceptable image quality.
Solution Approach 2:
The patent uses random masks as a form of data transformation that copies and modifies pixel information in a simplified manner. The masks act as templates that select or attenuate specific pixel values, creating a compressed representation that requires minimal processing power compared to traditional codecs.
3Device complexity
If lossy compression methods are used on mobile devices, then computational complexity is reduced, but material quality is lost
Solution Approach 1:
The patent applies random mask parameters that can be either binary (0 or 1) or grayscale (values between 0 and 1). These parameter changes create a transformed representation of the image that maintains quality information while enabling low-complexity processing. The masks effectively encode spatial frequency information in a way that preserves material quality without requiring complex computations.
Solution Approach 2:
Instead of applying complex transforms at the encoder and simple operations at the decoder (traditional approach), the patent inverts the complexity distribution by using simple mask multiplication at the encoder and more sophisticated inverse transformation at the decoder. This inversion allows mobile devices to use simple encoder operations while still achieving high quality through the decoder's capabilities.
Data Source
AI summary
Described are a method, apparatus and computer program product are provided for block modulating video and image compression. Different masks may be imposed on different blocks of an image and then the blocks summed to one block. In some embodiments, an encoder complexity of the encoder may be O(1). A method can be carried out that includes dividing an input image into a plurality of blocks; assigning different masks to each respective image block of the plurality of image blocks, the different masks having a size equal to that of each respective image block of the plurality of image blocks; modulating each respective image block of the plurality of blocks the assigned different masks to generate a plurality of modulated image blocks; summing the modulated plurality of blocks to at least one summed block; and quantizing the at least one summed block.