US Patent Application 17824372. LOSSY COMPRESSION TECHNOLOGY FOR SMALL IMAGE TILES simplified abstract
Contents
LOSSY COMPRESSION TECHNOLOGY FOR SMALL IMAGE TILES
Organization Name
Inventor(s)
Sreenivas Kothandaraman of Sammamish WA (US)
Abhishek R. Appu of El Dorado Hills CA (US)
Prosun Chatterjee of Bangalore (IN)
Mohamed Farook of Bangalore (IN)
LOSSY COMPRESSION TECHNOLOGY FOR SMALL IMAGE TILES - A simplified explanation of the abstract
This abstract first appeared for US patent application 17824372 titled 'LOSSY COMPRESSION TECHNOLOGY FOR SMALL IMAGE TILES
Simplified Explanation
- This patent application describes a technology for encoding and decoding images using a spatial transformation method. - The spatial transformation is applied to individual tiles within an image block, resulting in two sets of data: first sub-band data and second sub-band data. - The first sub-band data is used to predict residual data, while the second sub-band data is used to generate quantization data. - The residual data and quantization data together represent a lossy compressed version of the image. - The decoder technology can recover the first sub-band data from the residual data and scale it up to obtain the second sub-band data from the quantization data. - An inverse spatial transformation is then applied to the first and second sub-band data on a per tile basis, resulting in the reconstruction of the original image block.
Original Abstract Submitted
Methods, systems and apparatuses provide for encoder technology that conducts a spatial transformation on tiles in a block of an image, wherein the spatial transformation is conducted on a per tile basis and results in a first sub-band data and second sub-band data, predicts residual data from the first sub-band data, and generates quantization data from the second sub-band data, wherein the residual data and the quantization data represent a lossy compressed portion of the image. Additionally, decoder technology may recover first sub-band data from residual data, scale up to second sub-band data from quantization data, wherein the residual data and the quantization data represent a lossy compressed portion of an image, and conduct an inverse spatial transformation on the first sub-band data and the second sub-band data, wherein the inverse spatial transformation is conducted on a per tile basis and results in tiles in a block of the image.
