Author Login IJIRT Partner Network (IPN) Login

Enhancing Dark Image Exposure

  • Unique Paper ID: 164769
  • Volume: 10
  • Issue: 12
  • PageNo: 2933-2938
  • Abstract:
  • This project aims to enhance the visual quality of digital images through deep exposure correction using a frequency-domain approach. Leveraging deep learning techniques, a convolutional neural network (CNN) is trained to automatically correct exposure issues in diverse datasets. The novel aspect of this project lies in the incorporation of a frequency-domain loss function, utilizing Fourier transforms to address specific characteristics of image frequencies. The model is trained on a carefully curated dataset, and hyperparameter tuning ensures optimal performance. The exposure correction tool is seamlessly integrated into image processing pipelines, providing a user-friendly interface for individual or batch image correction. Performance metrics, including mean squared error, PSNR, and SSI, are employed to quantify the improvement in exposure correction. The robustness and generalization of the model are evaluated across varying scenes and lighting conditions. Extensive documentation details the entire process, from data collection to model architecture, enabling users to effectively utilize and understand the exposure correction tool. Iterative refinement based on user feedback ensures the adaptability and continual improvement of the proposed solution in real-world applications. This project successfully implemented a deep exposure correction model with a unique frequency-domain loss function. The convolutional neural network demonstrated effectiveness in automatically enhancing image exposure. The integration of Fourier transforms and careful dataset curation contributed to the model's robustness. The user-friendly exposure correction tool, backed by comprehensive documentation, offers practical utility. Continuous refinement based on user feedback ensures adaptability. This project contributes to improving image quality and holds potential for diverse applications in digital imagery.
Download article email to a friend

Cite This Article

  • ISSN: 2349-6002
  • Volume: 10
  • Issue: 12
  • PageNo: 2933-2938

Enhancing Dark Image Exposure

Related Articles

04 Jan, 2024

Enhancing Dark Image Exposure

Read More
Impact Factor
8.01 (Year 2024)
UGC Approved
Journal no 47859

Join Our IPN

IJIRT Partner Network

Submit your research paper and those of your network (friends, colleagues, or peers) through your IPN account, and receive 800 INR for each paper that gets published.

Join Now

Latest Publication

Archive

View More

Recent Conferences

NCSEM 2024

National Conference on Sustainable Engineering and Management - 2024 Last Date: 15th March 2024

Submit inquiry