• JPEG XT (ISO/IEC 18477) is an international standard which extends the legacy JPEG standard (ITU Recommendation T.81 | ISO/IEC 10918-1) for supporting high dynamic range (HDR) imaging in a totally backwards compatible method.
• JPEG XT Part 6 and 7 (ISO/IEC 18477-6 and 7) is implemented in DoubleBench’s JPEG XT software, which is based on libjpeg-turbo and compatible with JPEG XT Reference Software (ISO/IEC 18477-5). Our software supports lossy coding, integer coding, floating-point coding in both encoder and decoder.
• DoubleBench’s JPEG XT software provides a significant improvement in quality performance when comparing with JPEG XT Reference Software (ISO/IEC 18477-5).

• DoubleBench’s JPEG XT software also includes a built-in tone mapping operator which generates a better low dynamic range (LDR) base image.

DoubleBench’s JPEG XT Software LDR image	JPEG XT Reference Software LDR image

• Reference
  S. Choi, O. J. Kwon, D. Jang, and S. Choi, “Evaluation of Various Tone Mapping Operators for Backward Compatible JPEG Image Coding”, KSII Trans. on Internet and Information Systems, vol. 9, no. 9, pp. 3672-3684, Sep. 2015.
  S. Choi, O. J. Kwon, J. Lee, and Y. Kim, “A JPEG backward-compatible image coding scheme for high dynamic range images”, Digital Signal Processing, vol. 67, pp. 1-16, Aug. 2017.
  O. J. Kwon, S. Choi, and D. Shin, “Improvement of JPEG XT Floating-Point HDR Image Coding Using Region Adaptive Prediction”, IEEE Access, vol. 6, pp. 3321-3335, Feb. 2018.
• Patent
  Method and Apparatus for pre-processing of HDR image encoding apparatus based on a single quality value, Method for encoding HDR image based on a single quality value
        [KR]   Application Date: 2015.03.03 No.: 10-2015-0029769
                        Issue Date: 2021.08.12 No.: 10-2291585
  METHOD FOR ENCODING HIGH DYNAMIC RANGE IMAGE AND METHOD FOR DECODING HIGH DYNAMIC RANGE IMAGE
        [KR]   Application Date: 2016.02.19 No.: 10-2016-0019859
        [KR]   Application Date: 2016.05.26 No.: 10-2016-0064515
  IMAGE ENCODING AND DECODING METHODS, ENCODER AND DECODER USING THE METHODS
        [KR]   Application Date: 2017.02.03 No.: 10-2017-0015839
        [PCT] Application Date: 2017.02.17 No.: PCT/KR2017/001800
        [US]   Application Date: 2018.08.20 No.: 15/999,734
  Method and Apparatus for pre-processing of HDR image encoding apparatus based on a single quality value, Method for encoding HDR image based on a single quality value
        [KR]   Application Date: 2018.11.01 No.: 10-2018-0132672
        [PCT] Application Date: 2019.10.23 No.: PCT/KR2019/013933
        [US]   Application Date: 2021.02.05 No.: 17/266,546
                        Issue Date: 2022.12.13 No.: US11,528,491 B2

• JPEG 360 (ISO/IEC 19566-6) is an international standard which defines metadata enabling efficient way to exchange 360-degree images captured and shared through various capturing devices and applications.

Conventional way using JPEG images for 360-degree images	Standardized way using JPEG 360 images

• DoubleBench’s JPEG 360 software is the first application implementing JPEG 360 international standard in the world.
• JPEG to JPEG 360 converter: We support a conversion from the conventional JPEG file with customized metadata for 360-degree images to the JPEG 360 image file.
• JPEG and JPEG 360 image viewer: DoubleBench’s JPEG 360 image viewer is an application which parses and displays the JPEG 360 image file to extract metadata for 360-degree services.
  We provide standalone applications working on Microsoft Windows and Android devices. Also all web browsers supporting the WebGL™ can run our JPEG 360 image viewer.
  Converter&Viewer Demo
• JPEG 360 PTZ: generates JPEG 360 image in conjunction with PTZ cameras.
  

  JPEG 360 PTZ video

  JPEG 360 Phone: generates JPEG 360 image in conjunction with Phone cameras.

• Reference
  T. T. H. Uyen, O. J. Kwon, S. Choi, and I. Hussain, “Subjective Assessment of 360° Image Projection Formats”, IEEE Access, vol. 8, pp. 33588-33599, Feb. 2020.
  I. Hussain, O. J. Kwon, and S. Choi, “Evaluating the Coding Performance of 360° Image Projection Formats Using Objective Quality Metrics”, Symmetry, 13(1), 80, Jan. 2021.
  O. J. Kwon, J. Cho, and S. Choi, “A Center-to-Edge Progression for Equirectangular Projected 360° JPEG Images”, IEEE Access, vol. 9, pp. 6921-6929, Jan. 2021.
  I. Hussain and O. J. Kwon, “Evaluation of 360° Image Projection Formats; Comparing Format Conversion Distortion Using Objective Quality Metrics”, Journal of Imaging, 7, 137, Aug. 2021
  Ullah, F.; Kwon, O.-J.; Choi, S. Generation of a Panorama Compatible with the JPEG 360 International Standard Using a Single PTZ Camera. Appl. Sci. 2021, 11, 11019. https://doi.org/10.3390/app112211019
  Yaseen, O. J. Kwon, J. Lee, F. Ullah, S. Jamil, and J. S. Kim, “Automatic Sequential Stitching of High-Resolution Panorama for Android Devices Using Precapture Feature Detection and the Orientation Sensor”, Sensors, vol. 23, 879, Jan. 2023.
• Patent
  CENTER-TO-EDGE PROGRESSIVE IMAGE ENCODING/DECODING METHOD AND APPARATUS
        [KR]   Application Date: 2019.02.20 No.: 10-2019-0019853
        [PCT] Application Date: 2019.02.10 No.: PCT/KR2020/001831
        [US]   Application Date: 2021.02.05 No.: 17/266,562

• JPEG Snack (ISO/IEC 19566-8) is an international standard that enriches a representation of multiple media contents to facilitate sharing, editing, and presentation.

• Doublebench’s JPEG Snack software is the first application in the world to implement the JPEG Snack international standard.
• JPEG Snack  Encoder: DoubleBench’s JPEG Snack Encoder is an application that embeds multiple media content such as audio, videos, groups of images, and captions in the standard JPEG file
• JPEG Snack  Decoder: DoubleBench’s JPEG Snack Decoder is an application that decodes the JPEG snack file.
• JPEG Snack  Player: DoubleBench’s JPEG Snack Player provides the essence of the visual display of JPEG Snack content  to the users

JPEG Snack Player Demo

• Reference
  S. Jamil, O. J. Kwon, S. Choi, A. Kuzma, F. Ullah, Yaseen, and J. Lee, “Overview of JPEG Snack: A Novel International Standard for the Snack Culture”, IEEE Access, vol. 10, pp. 133402-133411, Dec. 2022.
  S. Jamil, O. J. Kwon, J. Lee, F. Ullah, Yaseen, and Afnan, “A Novel Multimedia Player for International Standard—JPEG Snack”, Journal of Imaging, 9, 58, Mar. 2023.
• Patent
  APPARATUS AND METHOD FOR STORING SNACK CULTURE CONTENTS
        [KR]   Application Date: 2020.10.12 No.: 10-2020-0130849
                        Issue Date: 2022.08.24 No.: 10-2437726
  Object oriented multimedia composition format for short-form contents
        [KR]   Application Date: 2021.10.28 No.: 10-2021-0145471
        [PCT] Application Date: 2022.10.04 No.: PCT/KR2022/014902
  CONTENT PROVIDING METHOD AND APPARATUS, AND CONTENT PLAYBACK METHOD
        [PCT] Application Date: 2021.09.06 No.: PCT/KR2021/012034
        [US]   Application Date: 2023.04.11 No.: 18/031,201

DoubleBench’s JPEG Privacy and Security Software for integrity.

• DoubleBench’s JPEG Privacy & Security is the first application in the world implementing the JPEG Systems Part-4 (ISO/IEC 19566-4) standard which provides integrity checking mechanisms for JPEG images. We support integrity checking functionality based on signature or watermark.

• Signature-based method

Support JPEG-1 (ISO/IEC 10918), JPEG XT (ISO/IEC 18477), and JPEG2000 (ISO/IEC 15444)

Embedding module	Verification module

• Watermark-based method

Support JPEG-1 (ISO/IEC 10918)

Embedding module	Verification module

Signature & Watermark based method Demo

• Phone version

DoubleBench’s JPEG Privacy and Security Software for Encryption.

• We provide a command-line tool for securing your images and protecting your privacy with ISO/IEC 19566-4 Standard.

DoubleBench’s encryption system, featuring AES-256-CBC, RSA-AES hybrid, JPEG compatibility, and CLI support.

• Key feature:
  • AES-256-CBC Standard Encryption: Our default encryption scheme uses the robust AES-256-CBC algorithm to provide a strong, internationally recognized layer of security for your images.
  • RSA-AES Hybrid Scheme: For an added layer of protection, our software offers a hybrid RSA-AES encryption option, leveraging asymmetric and symmetric algorithms for maximum security.
  • Full JPEG Compatibility: The encrypted output files remain fully compatible with the JPEG file format, ensuring seamless integration with existing image workflows.
  • Other than JPEG Files: The User can secure any file format. The generated output secured file will still be compatible with the standard. 
  • User-Friendly CLI: Designed for power users and developers, our command-line application offers granular control and easy automation for your security needs.

• Reference
  O. J. Kwon, S. Choi, and B. Lee, “A Watermark-Based Scheme for Authenticating JPEG Image Integrity”, IEEE Access, vol. 6, pp. 46194- 46205, Sep. 2018.
• Patent
  METHOD AND APPARATUS FOR VERIFYING INTEGRITY OF IMAGE BASED ON WATERMARK
        [KR]   Application Date: 2018.05.23 No.: 10-2018-0058185
                       Issued Date: 2024.02.08 No.: 10-2637177
        [PCT] Application Date: 2019.05.17 No.: PCT/KR2019/005890
        [US]   Application Date: 2020.11.20 No.: 17/057,249
                       Issued Date: 2024.02.27 No.: US11,915,336 B2

• DoubleBench’s JPEG-Payloader is a lightweight software that enables users to embed XML, JSON, CBOR or any types of files or metadata in his/her JPEG image. The software embeds the user data in JPEG Images according to the specifications of ISO/IEC 19566-5:2023 JPEG Universal Metadata Box Format (JUMBF) standard. User can use JPEG-Payloader for the following purposes:
  • Embed any kind of data/metadata using JUMBF box in JPEG images.
  • See what data/metadata is present in the JUMBF box in a JPEG image.
  • Extract any data/metadata from JUMBF box in a JPEG image.
  • Delete all or specific data/metadata from JUMBF box in a JPEG image.

JPEG Payloader Demo

• DoubleBench’s Scanned Book Aligner cuts content area(s) from a page of single-sided or double-sided books, automatically adjusts orientation, and removes shadows, which is useful for high-speed digitization of books.

Input	Output_1	Output_2

Demo video

• DoubleBench’s Cartoon Cut Segmenter segments individual cuts from a page of cartoon and orders them, which is useful for converting cartoon services from page-by-page style to cut-by-cut style.

Input	Output

Demo

• Reference
  O. J. Kwon, J. Lee, and S. Choi, “An Efficient Ordered-Cut Extracting Method for Scanned Cartoon Images”, Proceedings of ICCCA, KSCI, pp.75-78, Kanoa Resort Saipan, Saipan, USA, Aug. 2017.

• DoubleBench’s Multi-Exposure Image Fusion software generates an HDR image by acquiring images with different exposures in the same scene based on the fact that exposure change affects the change in the local luminance details, contrast, and colorfulness of a pixel.
• DoubleBench’s Multi-Exposure Image Fusion software allows the users to capture high-dynamic range images directly on digital cameras or smartphones, without using offline image-processing software.

Input	Output

Input	Output

• Reference
  S. Choi, O. J. Kwon, and J. Lee, “A Method for Fast Multi-Exposure Image Fusion”, IEEE Access, vol. 5, pp. 7371-7380, June 2017.

• DoubleBench’s Multi-Focus Image Fusion software generates an all-in-focus image by obtaining the relative focus measure between two images. This relative focus measure is based on the observation that the average filtered version of a well-focused region in an image shows a higher correlation to the corresponding defocused region in another image than the original well-focused version.
• DoubleBench’s Multi-Focus Image Fusion software allows the users to capture all-in-focus images directly from a digital camera or smartphone without using offline image processing software.

Input	Output

Input	Output

• Reference
  O. J. Kwon, S. Choi, D. Jang, and H. S. Pang, “All-in-focus imaging using average filter-based relative focus measure”, Digital Signal Processing, vol. 60, pp. 200-210, Jan. 2017.

Doublebench’s dynamic (spatio-temporal) Hand Gesture Recognition (HGR) module enables precise, real-time drone control and is highly customizable for other gesture-based control tasks such as Human–Computer Interaction (HCI).

This module is built on our proprietary benchmark dataset “DualExoGesture”, specifically designed and tailored for dual-hand activity recognition, ensuring robust performance in dynamic control scenarios.

Powered by a lightweight SqueezeNet–CNN hybrid model, it delivers real-time inference even on resource-constrained devices, making it ideal for field deployment.

Overall architecture of the proposed dynamic (spatio-temporal) Hand Gesture Recognition (HGR) module. The system captures dual-hand gesture sequences via an RGB camera, extracts spatio-temporal features using a lightweight SqueezeNet–CNN hybrid network, and maps the recognized gestures to predefined control commands for real-time drone operation. The modular design supports easy customization for other gesture-based control tasks such as HCI.

Key Features

• Real-Time Performance
  Optimized lightweight hybrid architecture ensures minimal latency.

• Dual-Hand Gesture Support
  Trained on our DualExoGesture dataset for high accuracy in complex dynamic gestures.

• Compact Gesture Set
  Offers a predefined set of intuitive gesture commands for quick mastery with minimal training.

• Customizable Commands
  Easily adapt and expand gesture-to-command mappings to suit your application.

• Ready-to-Use API
  Seamlessly integrate the module into your codebase with our well-documented API.

  Cross-Application Flexibility
  Suitable for drones, robotics, AR/VR systems, and general HCI.

• References

  • Yaseen, et al., “Vision-Based Gesture-Driven Drone Control in a Metaverse-Inspired 3D Simulation Environment,” Drones, vol. 9, no. 2, p. 92, Feb. 2025.

  • Yaseen, et al., “Evaluation of Benchmark Datasets and Deep Learning Models with Pre-Trained Weights for Vision-Based Dynamic Hand Gesture Recognition,” Applied Sciences, vol. 15, no. 11, p. 6045, Jun. 2025.

  • Yaseen, et al., “Next-gen Dynamic Hand Gesture Recognition: Mediapipe, Inception-v3 and LSTM-based Enhanced Deep Learning Model,” Electronics, vol. 13, no. 16, p. 3233, Aug. 2024.