c4 project batch submitted in MREC main campus ppt

project for B.TECH

1. MALLA REDDY ENGINEERING COLLEGE DEPARTMENT OF ELECTRONICS AND COMMUNICATION ENGINEERING A MAJOR PROJECT PRESENTATION ON DESIGN AND IMPLEMENTATION OF BILATERAL FILTER WITH EDGE PRESEVING ALGORITHM FOR NOISY IMAGE RESTORATION By Batch C4 M.Sindhu 20J41A04E7 M.Maddulety Yadav 20J41A04E5 B.Ankith Raj 20J41A04C7 S.Jagadeeswar Reddy 20J41A04H0 Under the guidance of Dr. Vasudeva Bevara (Assistant Professor)

2. CONTENTS Abstract Objective Literature survey Introduction Algorithm Working principle Block diagram Applications Results Conclusion Future scope References 2

3. ABSTRACT 3 This project presents a novel approach for restoring noisy images using a bilateral filter with an edge-preserving algorithm. The proposed method aims to effectively remove noise while preserving important edges and details in the image. The bilateral filter is utilized to smooth the image while retaining edge information, and the edge-preserving algorithm further enhances the preservation of significant features. Experimental results demonstrate the effectiveness of the proposed method in restoring noisy images, achieving superior performance compared to existing techniques. The combination of the bilateral filter and edge-preserving algorithm offers a promising solution for various applications in image restoration and enhancement. We provide visual and quantitative results on standard test images which show that this improvement is significant both visually and in terms of PSNR and SSIM.

4. OBJECTIVE To effectively reduce noise and to get a better reconstructed image as output, so that its suitable for many real-time applications. The objective of designing and implementing a bilateral filter with an edge-preserving algorithm for noisy image restoration is to develop a robust method for enhancing images corrupted by various types of noise while preserving important edges and structures. 4

5. LITERATURE REVIEW S.NO AUTHOR NAME & YEAR METHOD USED ADVANTAGES DISADVANTAGES 1 A.K. Jain(1989) Median filtering Simple to implement and much efficient Remove image details such as thin lines and corners 2 Scott E Umbaugh (1998) Mean filtering reduces the intensity variation between adjacent pixels. For small SNR’s it break up image edges & produce false noise edges 3 S.Balasubramanian (2008) Non-linear Cascade Filtering Enhanced image quality Computational complexity 4 Barwar Ferzo (2020) Wavelet based thresholding Automatically adopts to different peak widths Inefficient for computing geometrical features 5 Ademola E. Ilesanmi(2021) CNN(Deep learning) Extract image information easily Difficult to train image 5

6. 6 INTRODUCTION IMAGE - It is a visual representation formed by group of pixels NOISE- A random variation in an image where  y  is the observed noisy image,  x  is the clean image, and  n  represents noise. The purpose of noise reduction is to decrease the noise in natural images while minimizing the loss of original features and improving the peak-signal-to-noise ratio (PSNR). The major things which should be considered while denoising are: flat areas should be smooth, edges should be protected without blurring, textures should be preserved etc. y = x + n

7. A bilateral filter is a non-linear and noise-reducing smoothing filter for images. It replaces the intensity of each pixel with a weighted average of intensity values from nearby pixels. This weight can be based on a Gaussian distribution. Unlike traditional linear filters that treat all pixels equally, bilateral filtering considers both spatial and intensity information. Crucially, the weights depend not only on pixel distance (Euclidean distance) of pixels, but also on the differences in pixel values (radiometric differences - range differences, such as color intensity, depth distance, etc.). This preserves sharp edges. The filter combines a spatial Gaussian filter, which operates based on pixel distances, and an intensity Gaussian filter, which considers differences in pixel values. 7 BILATERAL FILTER EDGE PRESERVING Edge-preserving techniques are vital in image processing for maintaining sharp edges while reducing noise or performing other modifications. These methods, like bilateral filtering, median filtering, and guided filtering, distinguish between noise and actual edges, ensuring edge details remain intact during processing. This preservation is crucial in applications such as image denoising, enhancement, and edge detection across various fields like medical imaging, satellite imaging, photography, and computer vision.

8. BILATERAL FILTERING ALGORITHM 8 Bilateral filtering works by considering both spatial distance and intensity difference between pixels. For each pixel in the image, a weighted average is computed using a Gaussian function to measure spatial distance and another Gaussian function to measure intensity difference .

9. 9 EDGE PRESEVING ALGORITHM R( i,j )=Median(f( i,j ), b,d,e,g )

10. WORKING PRINCIPLE The design and implementation of a bilateral filter with an edge-preserving algorithm for noisy image restoration follows a sophisticated process aimed at achieving optimal noise reduction while preserving important image features. It works by applying a weighted average to each pixel in the image, taking into account both its spatial distance from neighboring pixels and the intensity This approach ensures that smoothing occurs while maintaining sharp edges. Additionally, the incorporation of an edge-preserving algorithm further refines the filtering process by identifying and preserving edges through the adjustment of filtering parameters. Trade-offs exist between computational speed and denoising accuracy, with applications in real-time image processing and resource-constrained environments. Evaluation involves metrics like PSNR and SSIM, ensuring robustness across different scenarios. 10

11. Block diagram 11 Generate texture map Generate block discontinuity map Calculate calculate Bilateral Filter Edge Preserving Algorithm Input Image Output Image

12. APPLICATIONS 12 Image Processing Computer Vision Medical Imaging Industrial Inspection Smartphone and Digtal Cameras Photography

13. RESULTS 13

14. 14 PSNR COMPARISION

15. conclusion In conclusion, the designed and implemented bilateral filter with an edge-preserving algorithm presents a robust and effective approach for restoring noisy images. Through experimental validation, we have demonstrated its superiority over conventional denoising methods in preserving important image features while effectively reducing noise, as evidenced by higher PSNR and SSIM values. 15

16. Future scope The future scope of approximate bilateral filtering for image denoising is promising, with several potential avenues for advancement and application: Deep Learning Integration: Combining traditional filtering methods with deep learning models could enhance denoising performance. Real-time Applications: Continued optimization of approximate bilateral filtering algorithms could enable real-time image processing in fields like augmented reality and medical imaging. Adaptive Filtering: Research focuses on developing adaptive techniques to dynamically adjust filter parameters based on image characteristics and noise levels. Multi-modal Data Fusion: Extending bilateral filtering to handle multi-modal data could enhance its utility in applications such as remote sensing and medical imaging. Application in Emerging Technologies: Approximate bilateral filtering may find new applications in computational photography, virtual reality, and digital entertainment for enhancing image quality and user experience. 16

17. REFERENCES 17 [1] P. K. R. Maddikunta , Q.-V. Pham, P. B, N. Deepa, K. Dev, “Industry 5.0: A Survey on Enabling Technologies and Potential Applications”, Journal of Industrial Information Integration, Vol.26, https://doi.org/10.1016/j.jii.2021.100257, n°3, pp. 1-31, Mar. 2022. [2] J. Nakamura, Image Sensors and Signal Processing for Digital Still Cameras, New York, NY, USA: Taylor & Francis, chap. 3, pp.55-90. [3] D. H. Shin, R. H. Park, S. Yang, J. H. Jung, “Block-Based Noise Estimation Using Adaptive Gaussian Filtering”, IEEE Transactions on Consumer Electronics, vol. 51, no. 1, DOI: 10.1109/TCE.2005.1405723, pp. 218-226, Febr . 2005. [4] C. Liu, W. T. Freeman, R. Szeliski , S. B. Kang, "Noise Estimation from a Single Image", in IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR), DOI: 10.1109/CVPR.2006.207, pp. 901-908, June 2006. [5] G. Chen, F. Zhu and P. A. Heng, "An Efficient Statistical Method for Image Noise Level Estimation", in IEEE International Conference on Computer Vision (ICCV), DOI: 10.1109/ICCV.2015.62, pp. 477-485, Dec. 2015.