Standard

You might also like

Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 12

Muhammad Mazhar

23-MS-EE-09
z
Image Compression using
Orthogonal Ramanujan
Wavelet Transform
z
Image Compression using Orthogonal
Ramanujan Wavelet Transform

 Published in: 2023 IEEE 3rd Applied Signal Processing


Conference (ASPCON)

 Date of Conference: November 2023

 Date Added to IEEE Xplore: January 2024

 Conference Location: India


z
Abstract

 A new image compression algorithm is introduced that combines the Discrete


Wavelet Transform with Orthogonal Ramanujan Sums.

 This method outperforms the standard JPEG compression technique in terms


of Mean Squared Error, Peak Signal-to-Noise Ratio, and Structural Similarity
Index at similar compression levels.

 Compared to JPEG, the new method reduces errors by up to 69.88%.The


proposed algorithm achieves a 24.2% improvement in PSNR, indicating better
image quality after compression.

 The new method also achieves a 27.3% improvement in SSIM, signifying


better preservation of the image's structural details after compression.
z

Problem Statement

 Image Compression using Orthogonal Ramanujan Wavelet


Transform" is to assess the effectiveness and practicality of ORWT
in image compression, including its performance, computational
efficiency, and suitability for various image datasets.
z
Introduction
•Image compression reduces the data needed to store and transmit digital images.

•It's crucial for efficient storage and transmission in areas like multimedia communication
and medical imaging.
•There are two main types: lossless (original image perfectly recovered) and lossy
(some information loss for smaller file size).
•Lossless compression is used for medical imaging and archives (uses techniques like
Huffman coding).
•Lossy compression is acceptable for applications like photography where minor quality
loss is tolerable (uses techniques like JPEG).
•This paper proposes a new lossy image compression algorithm combining Orthogonal
Ramanujan Sums (ORS) based Discrete Wavelet Transform (DWT) with JPEG
compression.
•The proposed method's performance is compared against standard JPEG
compression.
z
EXPERIMENTAL SETUP
 * Experiment used the standard "lenna.bmp" image: 512x512 RGB 8-bit (768 KiB).

 * Performance judged by:

 * Compression Ratio: Size of original image compared to compressed one.

 * Mean Squared Error (MSE): How different compressed image pixels are from
original (lower means better).

 * Peak Signal-to-Noise Ratio (PSNR): Ratio of maximum signal to noise


introduced by compression (higher means better).

 * Structural Similarity Index (SSIM): How similar the compressed image is to the
original (higher means better).

 * Software: Python with scikit-image and OpenCV libraries.


z
RESULT

• Proposed algorithm outperforms JPEG in image compression. This is evident


when comparing several key metrics:

• Mean Squared Error (MSE): Lower MSE indicates a smaller difference between the
original and compressed image. Our proposed method achieves consistently lower
MSE values.

• Peak Signal-to-Noise Ratio (PSNR): Higher PSNR indicates better image quality
after compression. The proposed method achieves higher PSNR compared to JPEG.

• Structural Similarity Index (SSIM): SSIM measures how similar the compressed
image is to the original, considering structure, contrast, and luminance. Higher SSIM
values indicate better similarity. The proposed method achieves higher SSIM values.
z
RESULT

•JPEG offers lower computational complexity compared to the proposed


method. This is an important consideration depending on your application's
needs.
•Time Complexity:
•Proposed Method: O(N³) - The dominant factor is the ORS-based DWT,
which involves matrix multiplications with this complexity.
•JPEG: O(N² log N) - The complexity is dominated by Huffman coding, which
scales with the number of unique symbols (pixels) in the image. (N represents
the image size in pixels)
•Trade-off: While the proposed method offers better compression performance,
it comes at the cost of higher computational time. This trade-off needs to be
considered based on your specific requirements. For applications where
compression speed is critical, JPEG might be a better choice.
z
z
z
z
Conclusion
•. Introduced a novel lossy image compression algorithm combining
ORS-based DWT with JPEG.
•Implemented the algorithm in Python for evaluation.
•Compared the proposed method against standard JPEG
compression.
•Chose JPEG for comparison as both are lossy, unlike PNG and
BMP (lossless formats).
•Lossless formats preserve all details (ideal for medical imaging) but
have larger file sizes.
•Proposed lossy algorithm achieves smaller file sizes by allowing
controlled information loss.

You might also like