Description:BACKGROUND
Field of Invention
[001] Embodiments of the present invention generally relate to an image compression technology, and particularly to a system and method for compressing ultra-high-resolution images using an enhanced discrete cosine transform (DCT) framework.
Description of Related Art
[002] With the growing demand for ultra-high-resolution imaging in fields such as medical diagnostics, satellite imaging, digital archiving, and high-end photography, there is a significant need for efficient image compression techniques that can handle large image sizes without compromising visual fidelity. Conventional image compression algorithms, such as JPEG, often rely on discrete cosine transform (DCT) based methods that are limited in performance when dealing with very high-resolution images, particularly in maintaining fine detail and minimizing artifacts.
[003] These conventional methods typically apply uniform quantization and fixed block sizes, which may not adapt well to local variations in image characteristics such as texture and contrast. Moreover, standard entropy encoding schemes may not fully exploit the statistical redundancy present in such images that leads to sub-optimal compression ratios. Furthermore, traditional compression techniques generally lack intelligent handling of perceptually significant regions, treating all parts of the image with the same compression logic, thereby affecting the overall visual quality. The computational efficiency of such systems is also inadequate for real-time applications, particularly when processing gigapixel-scale images.
[004] There is thus a need for an improved and advanced an improved image compression system that can administer the aforementioned limitations in a more efficient manner.
SUMMARY
[005] Embodiments in accordance with the present invention provide a system for compressing ultra-high-resolution images using an enhanced discrete cosine transform (DCT). The system comprises an input unit adapted to receive an ultra-high-resolution image, and a processing unit in communication with the input unit. The processing unit is configured to divide the received image into a plurality of blocks, apply a discrete cosine transform to each of the divided blocks to convert spatial domain data into frequency domain components, obtain quantized data by performing adaptive quantization on the frequency domain components, wherein the quantization is dynamically adjusted based on local image characteristics including texture and contrast, encode the obtained quantized data using an enhanced entropy encoding algorithm, and generate a compressed image output using an output unit based on the encoded quantized data. The generated compressed image is configured to preserve a visual quality and structural information of the received image.
[006] Embodiments in accordance with the present invention further provide a method for compressing ultra-high-resolution images using an enhanced discrete cosine transform (DCT). The method comprises receiving an image input using an input unit; dividing, by a processing unit, the received image into a plurality of blocks; applying a discrete cosine transform to each of the divided blocks to convert spatial domain data into frequency domain components; obtaining quantized data by performing adaptive quantization on the frequency domain components, wherein the quantization is dynamically adjusted based on local image characteristics including texture and contrast; encoding the quantized data using an enhanced entropy coding scheme; and generating a compressed image output based on the encoded quantized data, wherein the generated compressed image is configured to preserve a visual quality and structural information of the received image.
[007] Embodiments of the present invention may provide a number of advantages depending on their particular configuration. First, embodiments of the present application may provide a system for compressing ultra-high-resolution images while preserving essential image features and maintaining visual fidelity.
[008] Next, embodiments of the present application may provide a system that performs adaptive quantization based on local image characteristics, such as texture and contrast, thereby enabling intelligent data reduction without perceptual quality loss.
[009] Next, embodiments of the present application may provide a system for encoding quantized data using an enhanced entropy encoding algorithm, which results in higher compression efficiency and improved storage utilization.
[0010] Next, embodiments of the present application may provide a system that leverages region-based optimization techniques by analyzing statistical properties of image blocks prior to quantization, enhancing the overall compression process.
[0011] Next, embodiments of the present application may provide a system for minimizing chromatic distortion during compression by processing red, green, and blue color channels individually, resulting in improved color fidelity.
[0012] Next, embodiments of the present application may provide a system for dynamically selecting block sizes based on content analysis, allowing optimization of the trade-off between compression ratio and image detail retention. These and other advantages will be apparent from the present application of the embodiments described herein.
[0013] The preceding is a simplified summary to provide an understanding of some embodiments of the present invention. This summary is neither an extensive nor exhaustive overview of the present invention and its various embodiments. The summary presents selected concepts of the embodiments of the present invention in a simplified form as an introduction to the more detailed description presented below. As will be appreciated, other embodiments of the present invention are possible utilizing, alone or in combination, one or more of the features set forth above or described in detail below.
BRIEF DESCRIPTION OF THE DRAWINGS
[0014] The above and still further features and advantages of embodiments of the present invention will become apparent upon consideration of the following detailed description of embodiments thereof, especially when taken in conjunction with the accompanying drawings, and wherein:
[0015] FIG. 1 illustrates a schematic block diagram of a system for compressing ultra-high-resolution images, according to an embodiment of the present invention;
[0016] FIG. 2A depicts a target image, according to an embodiment of the present invention;
[0017] FIG. 2B depicts an comparison graph of Peak Signal-to-Noise Ratio (PSNR) versus Mean Squared Error (MSE), according to an embodiment of the present invention;
[0018] FIG. 2C depicts Mean Squared Error (MSE), according to an embodiment of the present invention;
[0019] FIG. 2D depicts Peak Signal-to-Noise Ratio (PSNR), according to an embodiment of the present invention;
[0020] FIG. 2E depicts a histogram of RED channel, according to an embodiment of the present invention;
[0021] FIG. 2F depicts a graph of compression ration Vs block size, according to an embodiment of the present invention;
[0022] FIG. 2G depicts a histogram of Blue channel, according to an embodiment of the present invention;
[0023] FIG. 2H depicts a histogram of green channel, according to an embodiment of the present invention; and
[0024] FIG. 3 depicts a flowchart of a method for real-time rice plant disease detection using an edge Artificial Intelligence, according to an embodiment of the present invention.
[0025] The headings used herein are for organizational purposes only and are not meant to be used to limit the scope of the description or the claims. As used throughout this application, the word "may" is used in a permissive sense (i.e., meaning having the potential to), rather than the mandatory sense (i.e., meaning must). Similarly, the words “include”, “including”, and “includes” mean including but not limited to. To facilitate understanding, like reference numerals have been used, where possible, to designate like elements common to the figures. Optional portions of the figures may be illustrated using dashed or dotted lines, unless the context of usage indicates otherwise.
DETAILED DESCRIPTION
[0026] The following description includes the preferred best mode of one embodiment of the present invention. It will be clear from this description of the invention that the invention is not limited to these illustrated embodiments but that the invention also includes a variety of modifications and embodiments thereto. Therefore, the present description should be seen as illustrative and not limiting. While the invention is susceptible to various modifications and alternative constructions, it should be understood, that there is no intention to limit the invention to the specific form disclosed, but, on the contrary, the invention is to cover all modifications, alternative constructions, and equivalents falling within the scope of the invention as defined in the claims.
[0027] In any embodiment described herein, the open-ended terms "comprising", "comprises”, and the like (which are synonymous with "including", "having” and "characterized by") may be replaced by the respective partially closed phrases "consisting essentially of", “consists essentially of", and the like or the respective closed phrases "consisting of", "consists of”, the like.
[0028] As used herein, the singular forms “a”, “an”, and “the” designate both the singular and the plural, unless expressly stated to designate the singular only.
[0029] FIG. 1 illustrates a schematic block diagram of a system 100 for compressing ultra-high-resolution images using an enhanced discrete cosine transform (DCT), according to an embodiment of the present invention. In an embodiment, the system 100 may be adapted to receive an ultra-high-resolution image, process the image through a series of compression stages, and output a compressed version while preserving visual fidelity and structural information. Moreover, the system 100 may dynamically analyze image content to optimize compression parameters in real-time. Furthermore, the system 100 may apply adaptive quantization and enhanced entropy encoding techniques based on local image characteristics. The system 100 may utilize a combination of hardware acceleration and algorithmic optimization to enable high-throughput and low-latency image compression, according to an embodiment of the present invention.
[0030] According to the embodiments of the present invention, the system 100 may incorporate non-limiting hardware components to enhance the processing speed and efficiency such as the system 100 may comprise an input unit 102, a processing unit 104, an output unit 106, and a graphical processing unit 108.
[0031] In an embodiment of the present invention, the input unit 102 may be adapted to receive ultra-high-resolution images for compression. The input unit 102 may be, but not limited to image sensors, digital cameras, satellite feeds, scanners, external storage devices, a computing device and so forth. Embodiments of the present invention are intended to include or otherwise cover any type of the input unit 102 including known, related art, and/or later developed technologies.
[0032] In an embodiment of the present invention, the input unit 102 may be adapted to receive the ultra-high-resolution images from a variety of input sources, including but not limited to satellite imaging, astronomical datasets, medical imaging platforms, an internet, and so forth. Embodiments of the present invention are intended to include or otherwise cover any type of the input sources including known, related art, and/or later developed technologies. In an embodiment of the present invention, the input unit 102 may preprocess the received image to ensure format compatibility and data integrity before forwarding the data to the processing unit 104. The preprocessing may optionally include image normalization, color space conversion, and segmentation into processable blocks based on content analysis.
[0033] The processing unit 104 may comprise programming modules to execute the computer-executable instructions, according to an embodiment of the present invention. The programming modules may include, but are not limited to, a block division module, a DCT transformation module, an adaptive quantization module, an entropy encoding module, and an output control module.
[0034] In an embodiment of the present invention, the block division module may be configured to segment the input image into a plurality of blocks of varying or fixed sizes, based on content analysis.
[0035] In an embodiment of the present invention, the DCT transformation module may be configured to convert each block from the spatial domain to the frequency domain using a discrete cosine transform. This transformation facilitates the separation of image content into frequency components, wherein lower frequencies represent general structural and tonal information, and higher frequencies represent finer details and noise. By concentrating most of the visually significant data into fewer coefficients, the DCT transformation enables more effective compression. The transformation may be applied independently to each image block, allowing for localized frequency analysis. In certain embodiments, the DCT may be implemented using fast algorithms optimized for parallel computation, such as those compatible with graphical processing units (GPUs), to accelerate processing of ultra-high-resolution images. Additionally, the transformation may be extended to apply on individual color channels or in a luminance-chrominance space to further enhance perceptual compression efficiency.
[0036] In an embodiment of the present invention, the adaptive quantization module may be configured to analyze local features such as texture complexity and contrast to apply variable quantization levels across different image regions. The adaptive quantization module may compute statistical metrics including pixel intensity variance, edge density, or entropy within each block to assess its visual significance. Regions with high texture or contrast, such as areas containing fine details or important structural patterns that may be assigned finer quantization scales to preserve fidelity, while smoother or uniform areas may be quantized more aggressively to reduce data volume. This adaptive approach mitigates artifacts such as blurring or blocking and enhances the overall perceptual quality of the compressed image. In some embodiments, the quantization thresholds may be dynamically adjusted based on global image characteristics or compression goals, enabling a context-aware trade-off between image quality and compression ratio. In an embodiment of the present invention, the entropy encoding module may be configured to utilize enhanced encoding techniques to minimize data redundancy in the quantized frequency domain.
[0037] In an embodiment of the present invention, the output control module may be configured to compile the encoded data into a compressed image format compatible with standard or proprietary output systems.
[0038] In some embodiments of the present invention, the processing unit 104 may also include a region-based analysis module configured to identify regions of interest based on statistical patterns, visual salience, or color/intensity variance. The region-based analysis module may optimize compression strategies by applying higher fidelity encoding to visually important areas and more aggressive compression to less critical regions. These optimizations may collectively enhance compression efficiency while retaining image quality in perceptually significant areas.
[0039] In an embodiment of the present invention, the output unit 106 may be adapted to generate and transmit the compressed image data produced by the processing unit 104. The output unit 106 may be configured to support various image formats, file containers, or data streaming protocols. Additionally, the output unit 106 may be operable to store the compressed image locally, transmit it to a remote system, or display a preview of the compressed output in real-time. The output unit 106 may further include optional functionality to assess and report the compression quality metrics, such as Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), Mean Squared Error (MSE), and so forth.
[0040] In an embodiment of the present invention, the graphical processing unit (GPU) 108 may be adapted to accelerate computation-intensive tasks associated with image compression, such as matrix transformations, pixel-level analysis, and parallel block processing. The GPU 108 may operate in conjunction with the processing unit 104 to execute parallelizable components of the compression algorithm, thereby improving throughput and reducing latency. The GPU 108 may also handle real-time visualization or encoding of image frames when the system 100 is employed in applications involving live image capture or streaming.
[0041] FIG. 2A depicts a target image 200, according to an embodiment of the present invention. In an embodiment of the present invention, the target image 200 may be sourced from the NASA James Webb Space Telescope (JWST) dataset and may represent a high-resolution astronomical image used as an input for testing the system 100. Upon receiving the target image 200, the system 100 may initiate the compression process by dividing the image into multiple blocks.
[0042] Based on the content characteristics of each region within the image, the system 100 may dynamically determine appropriate block sizes to balance between compression efficiency and preservation of fine details. The system 100 may then convert each image block from the spatial domain to the frequency domain using a discrete cosine transform, that may help in separating essential structural components from less perceptible data. A block-based methodology with a sliding window was employed to address computational demands when processing such ultra-high-resolution images. After transformation, the frequency components may undergo quantization, wherein the system 100 may adaptively adjust quantization parameters in response to localized features such as contrast intensity and texture complexity. This ensures that critical visual and structural elements within the image such as star formations or deep-space nebulae, may be retained with high fidelity, while less important regions may be compressed more aggressively to optimize the overall data size. Experimental evaluation using the MATLAB R2021a environment demonstrated performance of the system 100 on the JWST dataset. The system 100 achieved a significant compression ratio of 12.68, reducing image sizes from approximately 152MB to 12MB, while maintaining a high average PSNR of 39.67 dB and a Structural Similarity Index (SSIM) of 0.9906, indicating effective preservation of structural integrity.
[0043] The quantized data may then be encoded using an enhanced entropy coding algorithm. The encoding may enable compact representation of the image while reducing redundancy. Once encoded, the compressed image may be generated, transmitted and/or stored via the output unit 106, with minimal perceptual loss and high structural accuracy. Subjective and objective evaluations, including Mean Opinion Score (MOS), Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), Mean Squared Error (MSE), and Normalized MSE (NMSE), were employed to assess visual quality and compression performance. Additionally, color analysis indicated some loss of chromatic information particularly in the blue channel with an average Delta E of 4.8247. The study may further highlight that variations in block size within DCT-based compression directly impact the balance between computational efficiency and detail preservation. In this manner, the system 100 may enable efficient compression of ultra-high-resolution astronomical imagery like target image 200, while preserving the scientific and visual integrity of the original data.
[0044] FIG. 2B depicts a comparison graph of Peak Signal-to-Noise Ratio (PSNR) versus Mean Squared Error (MSE), according to an embodiment of the present invention. In an embodiment of the present invention, the inverse relationship between the image quality (PSNR) and the error (MSE), may demonstrate how various compression settings impact reconstructed image fidelity.
[0045] FIG. 2C depicts a Mean Squared Error (MSE) histogram, according to an embodiment of the present invention. In an embodiment of the present invention, the histogram may display a distribution of MSE values across image blocks, providing insight into localized error behavior during compression.
[0046] FIG. 2D depicts a Peak Signal-to-Noise Ratio (PSNR) histogram, according to an embodiment of the present invention. In an embodiment of the present invention, the histogram may reveal a variability of PSNR values across image segments, aiding in the assessment of structural integrity and quality retention.
[0047] FIG. 2E depicts a histogram of the Red channel, according to an embodiment of the present invention. In an embodiment of the present invention, this histogram may show the intensity distribution of red pixels in the compressed image, helping analyze color preservation and distortion.
[0048] FIG. 2F depicts a graph of compression ratio versus block size, according to an embodiment of the present invention. In an embodiment of the present invention, the graph may demonstrates how changes in block size influence compression efficiency and detail preservation, with trade-offs between higher ratios and visual quality.
[0049] FIG. 2G depicts a histogram of the Blue channel, according to an embodiment of the present invention. In an embodiment of the present invention, this histogram may highlight intensity variation in the blue spectrum, where slight losses in color fidelity were observed during compression.
[0050] FIG. 2H depicts a histogram of the Green channel, according to an embodiment of the present invention. In an embodiment of the present invention, this histogram may illustrate the green pixel distribution post-compression, supporting the evaluation of chromatic consistency across color channels.
[0051] FIG. 3 depicts a flowchart of a method 300 for compressing ultra-high-resolution images using an enhanced discrete cosine transform (DCT), according to an embodiment of the present invention. The method may be performed by using the system 100 and may be implemented in hardware, software, or a combination thereof.
[0052] At step 302, the system 100 may receive the ultra-high-resolution image via the input unit 102. The input image may be obtained from a variety of sources, such as, the satellite imaging, the astronomical datasets, the medical imaging platforms, the Internet, and so forth. The receive image may be fed into the processing pipeline using the system 100 for compression.
[0053] At step 304, the system 100 may divide the received image into the plurality of blocks. These blocks may be of fixed or variable size and serve as the basic units for applying frequency transformation and further compression. The system 100 may identify the visually important regions within the image and apply the lossless compression to such regions.
[0054] At step 306, the system 100 may apply the discrete cosine transform to each of the divided blocks to convert the spatial domain data into the frequency domain components. This transformation may enable energy compaction by allowing significant information to be concentrated in a smaller number of coefficients.
[0055] At step 308, the system 100 may obtain the quantized data by performing adaptive quantization on the frequency domain components. The quantization parameters may be dynamically adjusted based on the local image characteristics such as texture complexity and/or contrast for optimizing data reduction while preserving perceptual quality.
[0056] At step 310, the system 100 may encode the quantized data using the enhanced entropy coding algorithm. This step may involve techniques such as run-length encoding, Huffman coding, arithmetic coding, and so forth, to maximize compression efficiency for the statistical characteristics of the high-resolution image data.
[0057] At step 312, the system 100 may generate the compressed image output using the output unit 106. The compressed image may retain the key visual and the structural information. The system 100 may evaluate the generated compressed image using quality metrics such as Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), and Mean Squared Error (MSE).
[0058] While the invention has been described in connection with what is presently considered to be the most practical and various embodiments, it is to be understood that the invention is not to be limited to the disclosed embodiments, but on the contrary, is intended to cover various modifications and equivalent arrangements included within the scope of the appended claims.
[0059] This written description uses examples to disclose the invention, including the best mode, and also to enable any person skilled in the art to practice the invention, including making and using any devices or systems and performing any incorporated methods. The patentable scope of the invention is defined in the claims, and may include other examples that occur to those skilled in the art. Such other examples are intended to be within the scope of the claims if they have structural elements that do not differ from the literal language of the claims, or if they include equivalent structural elements within substantial differences from the literal languages of the claims. , Claims:CLAIMS
I/We Claim:
1. A system (100) for compressing ultra-high-resolution images using an enhanced discrete cosine transform (DCT), the system (100) comprising:
an input unit (102) adapted to receive an ultra-high-resolution image; and
a processing unit (104) in communication with the input unit (102), characterized in that the processing unit (104) is configured to:
divide the received image into a plurality of blocks;
apply a discrete cosine transform to each of the divided blocks to convert spatial domain data into frequency domain components;
obtain quantized data by performing adaptive quantization on the frequency domain components, wherein the quantization is dynamically adjusted based on local image characteristics including texture and contrast;
encode the obtained quantized data using an enhanced entropy encoding algorithm; and
generate a compressed image output, using an output unit (106) based on the encoded quantized data, wherein the generated compressed image is configured to preserve a visual quality and structural information of the received image.

2. The system (100) as claimed in claim 1, wherein the processing unit (104) is configured to analyze the statistical properties of the blocks for performing region-based optimization prior to quantization.

3. The system (100) as claimed in claim 1, wherein the input unit (102) and the output unit are operably connected to a graphical processing unit (GPU) (108) to support real-time image compression.

4. The system (100) as claimed in claim 1, wherein the processing unit (104) is configured to process individual red, green, and blue color channels to minimize chromatic distortion during compression.

5. The system (100) as claimed in claim 1, wherein the processing unit (104) is configured to dynamically select block sizes based on content analysis to optimize between compression ratio and image detail retention.

6. A method for compressing ultra-high-resolution images using an enhanced discrete cosine transform (DCT), the method comprising:
receiving an image input using an input unit (102);
dividing, by a processing unit (104), the received image into a plurality of blocks;
applying a discrete cosine transform to each of the divided blocks to convert spatial domain data into frequency domain components;
obtaining quantized data by performing adaptive quantization on the frequency domain components, wherein the quantization is dynamically adjusted based on local image characteristics including texture and contrast;
encoding the quantized data using an enhanced entropy coding scheme; and
generating a compressed image output based on the encoded quantized data, wherein the generated compressed image is configured to preserve a visual quality and structural information of the received image.

7. The method as claimed in claim 6, wherein the adaptive quantization is performed by analysing pixel intensity variation within each of the divided blocks to determine optimal quantization levels.

8. The method as claimed in claim 6, wherein the blocks with analogous statistical properties are grouped to apply region-based DCT optimization for improved compression efficiency.

9. The method as claimed in claim 6, comprising a step of identifying visually important regions within the image and applying lossless compression to such regions.

10. The method as claimed in claim 6, further comprising a step of evaluating the quality of the compressed image using metrics selected from Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), Mean Squared Error (MSE), or a combination thereof.
Date: April 22, 2025
Place: Noida

Nainsi Rastogi
Patent Agent (IN/PA-2372)
Agent for the Applicant