Scribd
Upload
9 views2 pages

2 Report

Uploaded by

cityofsuccess001
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
9 views2 pages

2 Report

Uploaded by

cityofsuccess001
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

2.1.

3 Image Deblurring: Image deblurring techniques aim to restore images that are blurred due to

factors like motion, camera shake, or lens aberrations (Kundur and Hatzinakos, 1996). The concept of image

deblurring is demonstrated in figure 2.3.

Figure 2.3: Image Deblurring

Image Deblurring Techniques

The following are the techniques that are used to represent image deblurring process

i. Deconvolution: This involves reversing the blurring process using the point spread function (PSF)

(Kundur and Hatzinakos, 1996).

ii. Blind deconvolution: This involves estimating the PSF and then deconvolving the image (Levin et

al., 2011).

iii. Deep learning-based methods: These methods use neural networks to learn the deblurring process

(Sun et al., 2015).

2.1.4 Image Enhancement: Image enhancement techniques aim to improve the visual quality of images

(Liu et al., 2018). It can be expressed in figure 2.4

Image Enhancement Techniques

Techniques for image enhancement can be stated as follows:

i. Histogram equalization: This involves adjusting the image contrast to make it more visually

appealing (Gonzalez and Woods, 2018).

ii. Contrast stretching: This involves stretching the image contrast to make it more visually appealing

(Jain, 2010).
5
2.5 Practical Examples of Investigation.

The practical examples of investigation is given as:

2.5.1 Image Segmentation using Deep Learning for Low-Quality Medical Images with Complex

Backgrounds.

1. Techniques:

i. Convolutional Neural Networks (CNNs): This is used for image segmentation to separate objects

from complex backgrounds.

ii. U-Net Architecture: This is a type of CNN specifically designed for image segmentation tasks.

iii. Data Augmentation: This is a technique used to artificially increase the size of the training dataset,

such as rotation, scaling, and flipping.

2. Investigation

Ronnerberger et al. (2015) investigated the use of deep learning techniques for image segmentation

in low-quality medical images with complex backgrounds. They use a U-Net architecture to segment cells

in microscope images with varying levels of quality and background complexity.

3. Results

The authors achieved high accuracy in segmenting cells in low-quality images with complex

backgrounds, outperforming traditional image processing techniques.

4. Evaluation

This study demonstrates the effectiveness of deep learning techniques in handling low-quality

images and complex backgrounds. The use of data augmentation techniques and U-Net architecture

can be applied to various image segmentation tasks, including medical image analysis.

You might also like