International Journal of Computer Science & Network Security

International Journal of Computer Science & Network Security (국제컴퓨터통신보호논문지학회)
권호 표지
DOI QR코드

DOI QR Code

Multi-type Image Noise Classification by Using Deep Learning

  • Waqar Ahmed (Center of Excellence for Robotics, Artificial Intelligence and Blockchain, Sukkur IBA University) ;
  • Zahid Hussain Khand (Center of Excellence for Robotics, Artificial Intelligence and Blockchain, Sukkur IBA University) ;
  • Sajid Khan (Center of Excellence for Robotics, Artificial Intelligence and Blockchain, Sukkur IBA University) ;
  • Ghulam Mujtaba (Center of Excellence for Robotics, Artificial Intelligence and Blockchain, Sukkur IBA University) ;
  • Muhammad Asif Khan (Department of Electrical Engineeing, Sukkur IBA University) ;
  • Ahmad Waqas (Center of Excellence for Robotics, Artificial Intelligence and Blockchain, Sukkur IBA University)
  • Received : 2024.07.05
  • Published : 2024.07.30
Download PDF
⟨ Previous Next ⟩

Abstract

Image noise classification is a classical problem in the field of image processing, machine learning, deep learning and computer vision. In this paper, image noise classification is performed using deep learning. Keras deep learning library of TensorFlow is used for this purpose. 6900 images images are selected from the Kaggle database for the classification purpose. Dataset for labeled noisy images of multiple type was generated with the help of Matlab from a dataset of non-noisy images. Labeled dataset comprised of Salt & Pepper, Gaussian and Sinusoidal noise. Different training and tests sets were partitioned to train and test the model for image classification. In deep neural networks CNN (Convolutional Neural Network) is used due to its in-depth and hidden patterns and features learning in the images to be classified. This deep learning of features and patterns in images make CNN outperform the other classical methods in many classification problems.

Keywords

References

  1. F. Russo, Edge detection in noisy images using fuzzy reasoning. IEEE Trans. Instr. and Meas. 47(5), 1102-1105 (1998) https://doi.org/10.1109/19.746564
  2. Khan, S., Lee, D. H., Khan, M. A., Gilal, A. R., Iqbal, J., & Waqas, A. (2020). Efficient and improved edge detection via a hysteresis thresholding method. Current Science (00113891), 118(6).
  3. Bustince, H., Barrenechea, E., Pagola, M., & Fernandez, J. (2009). Interval-valued fuzzy sets constructed from matrices: Application to edge detection. Fuzzy Sets and systems, 160(13), 1819-1840. https://doi.org/10.1016/j.fss.2008.08.005
  4. Khan, S., Lee, D. H., Khan, M. A., Gilal, A. R., & Mujtaba, G. (2019). Efficient edge-based image interpolation method using neighboring slope information. IEEE Access, 7, 133539-133548. https://doi.org/10.1109/ACCESS.2019.2942004
  5. Khan, S., & Lee, D. (2015). Efficient deinterlacing method using simple edge slope tracing. Optical Engineering, 54(10), 103108.
  6. Khan, S., Lee, D. H., Khan, M. A., Siddiqui, M. F., Zafar, R. F., Memon, K. H., & Mujtaba, G. (2020). Image Interpolation via Gradient Correlation-Based Edge Direction Estimation. Scientific Programming, 2020.
  7. Khan, S., Lee, D. H., Khan, A., Waqas, A., Gilal, A. R., & Khand, Z. H. (2020). A Digital Camera-Based Rotation-Invariant Fingerprint Verification Method. Scientific Programming, 2020.
  8. Khan, S., Waqas, A., Khan, M. A., & Ahmad, A. W. (2018). A camera-based fingerprint registration and verification method. International Journal of Computer Science and Network Security, 18(11), 26-31.
  9. Li, Q., Cai, W., Wang, X., Zhou, Y., Feng, D. D., & Chen, M. (2014, December). Medical image classification with convolutional neural network. In 2014 13th International Conference on Control Automation Robotics & Vision (ICARCV) (pp. 844-848). IEEE.
  10. E. Abreu, M. Lightstone, S.K. Mitra, K. Arakawa, A new efficient approach for the removal of impulse noise from highly corrupted image. IEEE Trans. Image Process. 5(6), 1012-1025 (1996) https://doi.org/10.1109/83.503916
  11. H.L. Eng, K.K. Ma, Noise adaptive soft-switching median filter. IEEE Trans. Image Process. 10(2), 242-251 (2001) https://doi.org/10.1109/83.902289
  12. Wang, T., Qiu J., Fu, S., Ji, W.: Distributed fuzzy H∞ filtering for nonlinear multirate networked double-layer industrial processes. IEEE Trans. Industrial Electronics 64(6), 5203-5211 (2017) https://doi.org/10.1109/TIE.2016.2622234
  13. P. Civicioglu, Using uncorrupted neighborhoods of the pixels for impulsive noise suppression with ANFIS. IEEE Trans. Image Process. 16(3), 759-773 (2007) https://doi.org/10.1109/TIP.2007.891067
  14. Khan, S., & Lee, D. H. (2017). An adaptive dynamically weighted median filter for impulse noise removal. EURASIP Journal on Advances in Signal Processing, 2017(1), 67.
  15. Lee, D. H. (2012). An Edge-Based Adaptive Method for Removing High-Density Impulsive Noise from an Image While Preserving Edges. ETRI Journal, 34(4), 564-571. https://doi.org/10.4218/etrij.12.0111.0392
  16. Majeeth, S. S., & Babu, C. N. K. (2019). Gaussian noise removal in an image using fast guided filter and its method noise thresholding in medical healthcare application. Journal of medical systems, 43(8), 280.
  17. Varghese, J., Subhash, S., Subramaniam, K., & Sridhar, K. P. (2020). Adaptive Gaussian notch filter for removing periodic noise from digital images. IET Image Processing, 14(8), 1529-1538. https://doi.org/10.1049/iet-ipr.2018.5707
  18. Khaw, Hui Ying & Soon, Foo Chong & Chuah, Joon Huang & Chow, Chee Onn. (2017). Image Noise Types Recognition Using Convolutional Neural Network with Principal Components Analysis. IET Image Processing. 11. 10.1049/iet-ipr.2017.0374.
  19. Roy, Sudipta & Ahmed, Mahtab & Akhand, M. A. H.. (2018). Noisy image classification using hybrid deep learning methods. Journal of Information and Communication Technology. 17. 233-269. 10.32890/jict2018.17.2.4.
  20. Gorkem Algan & IIkay Ulusoy (2020). Image Classification with Deep Learning in the Presence of Noisy Labels: A Survey arXiv:1912.05170v2 [cs.LG]
  21. Koziarski, Michal & Cyganek, Boguslaw. (2017). Image recognition with deep neural networks in presence of noise - Dealing with and taking advantage of distortions. Integrated Computer-Aided Engineering. 24. 1-13. 10.3233/ICA-170551.
  22. Natual Images | Kaggle, https://www.kaggle.com/prasunroy/natural-images, last accessed: 20/07/2020.
  23. Add noise to image- Matlab imnoise https://www.mathworks.com/help/images/ref/imnoise.html, last accessed: 20/07/2020.