JPEG2000: Still Image CompressionIntroductionMotivationOverviewSlide 5Progress & GoalsJPEG2000: Still Image CompressionJPEG2000: Still Image CompressionShyh-Jye NiShyh-Jye NiShoaib JunaidShoaib JunaidIntroductionIntroductionOver the past few years, digital imaging has gone Over the past few years, digital imaging has gone from a high-end niche capability to a ubiquitous from a high-end niche capability to a ubiquitous mainstream application, with the widespread use mainstream application, with the widespread use of digital pictures greatly enriching the content of of digital pictures greatly enriching the content of both commercial and personal communications. both commercial and personal communications. With millions of companies and consumers using With millions of companies and consumers using digital images in both print and on-line digital images in both print and on-line environments, the need has emerged for environments, the need has emerged for developing a new, comprehensive, flexible and developing a new, comprehensive, flexible and universally deployable file format for digital universally deployable file format for digital images.images.MotivationMotivationImproved Compression EfficiencyImproved Compression EfficiencyRicher Content and Capacity for Alternate Richer Content and Capacity for Alternate Color SpacesColor SpacesSupport for Flexible "Level of Interest" Support for Flexible "Level of Interest" AccessAccessConservation of BandwidthConservation of BandwidthOverviewOverviewPre-Process:Pre-Process:–Partition the input image into rectangular and non-overlapping tiles of equal size Partition the input image into rectangular and non-overlapping tiles of equal size (except possibly for those tiles at the image borders). The tile size is arbitrary and (except possibly for those tiles at the image borders). The tile size is arbitrary and can be as large as the original image itself or as small as a single pixel. can be as large as the original image itself or as small as a single pixel. –Next, unsigned sample values in each component are level shifted (DC offset) by Next, unsigned sample values in each component are level shifted (DC offset) by subtracting a fixed value of 2subtracting a fixed value of 2B_1B_1 from each sample to make its value symmetric from each sample to make its value symmetric around zero.around zero.–Finally, the level-shifted values can be subjected to a forward point-wise Finally, the level-shifted values can be subjected to a forward point-wise intercomponent transformation to decorrelate the color data.intercomponent transformation to decorrelate the color data.Pre-ProcessingDiscrete Wavelet Transform (DWT)Uniform Quantizer with DeadzoneTier-1 CodingTier-2 CodingOriginal Image DataCompressed Image DataJPEG 2000 fundamental building blocksOverviewOverviewDWT:DWT:–The DCT has been replaced by DWT, because the DWT has multi-resolution image The DCT has been replaced by DWT, because the DWT has multi-resolution image representation inherent to it. The use of Integer DWT filters allows for both lossless representation inherent to it. The use of Integer DWT filters allows for both lossless and lossy compression within a single bit stream.and lossy compression within a single bit stream.Quantizer:Quantizer:–Quantization is the element of lossy compression systems responsible for reducing Quantization is the element of lossy compression systems responsible for reducing the precision of data in order to make them more compressible.the precision of data in order to make them more compressible.Tier-1:Tier-1:–The arithmetic coding of the bitplane data is referred to as tier-1 (T1) coding.The arithmetic coding of the bitplane data is referred to as tier-1 (T1) coding.Tier-2:Tier-2:–The compressed sub-bitplane coding passes can be aggregated into larger units The compressed sub-bitplane coding passes can be aggregated into larger units named packets. This process of packetization along with its supporting syntax is named packets. This process of packetization along with its supporting syntax is often referred to as tier-2 (T2) coding.often referred to as tier-2 (T2) coding.Progress & GoalsProgress & GoalsProgress:Progress:–1. pre-processing1. pre-processing–2. DWT2. DWTGoals:Goals:–Minimum goal is to achieve the first 3 blocks, pre-processing, Minimum goal is to achieve the first 3 blocks, pre-processing, DWT and uniform quantizier with deadzone.DWT and uniform quantizier with
View Full Document