2024 International Conference on Information Technology, Electronics and Intelligent Communication Systems (ICITEICS)

2024 International Conference on Information Technology, Electronics and Intelligent Communication Systems (ICITEICS).

[Audio] Students today we will be discussing the characteristics and challenges of underwater images. Our oceans and underwater environments are vast and filled with beauty and mysteries but capturing clear and high-quality images in such an environment can be a daunting task. Underwater images possess unique characteristics that make them different from images taken on land. The main challenge in underwater image enhancement is the loss of color and contrast due to the absorption and scattering of light in the water. This can result in blurry and washed out images making it difficult to fully appreciate the beauty of the underwater world. Another challenge is the presence of particles and debris in the water which can cause distortions and obstructions in the image and make it challenging for researchers and scientists to analyze and interpret them accurately. Throughout this conference we will explore various techniques and technologies that can help overcome these challenges and enhance underwater images. Let's dive deep into the world of underwater imaging and discover the exciting possibilities that lie beneath the surface..

[Audio] Continuing to advance in technology has heightened our need for resources. The ocean covers 71% of our planet's surface making it a valuable source of resources. However this task presents its own set of challenges. That's why we are utilizing autonomous underwater vehicles or A-U-V-s for this important work. A-U-Vs offer a safer and more efficient method for exploring and developing the ocean as well as lakes and rivers. Visual information plays a crucial role in this process as it allows us to understand and detect our underwater surroundings. However obtaining high-quality images in this environment is difficult due to the unpredictable nature of the water and the effects of light absorption and scattering. In this presentation we will discuss the significance of visual information and the obstacles we face in obtaining it in underwater environments. So let us take a dive into how A-U-Vs are transforming our exploration and development of the vast ocean..

[Audio] Continuing our discussion on the applications of underwater images we will now explore some of the most significant areas where these images are utilized. These include marine biology research underwater archaeology inspection and maintenance fisheries management and oceanography. Underwater images greatly enhance our understanding of the world beneath the surface. They also have practical uses in environmental monitoring underwater photography and filmmaking and even in tourism and recreation. Furthermore they have crucial applications in military and defense as well as in public education such as in aquariums. Let's also not overlook their role in underwater infrastructure planning exploration and adventure. The possibilities and benefits of using underwater images are limitless and we will continue to delve deeper into these areas in our upcoming sessions..

[Audio] Continuing our exploration of challenges in obtaining quality underwater video images slide number 7 reveals common issues that can impact the footage. These include color bias atomization phenomenon blur low contrast color distortion more noise unclear details and limited visual range. These factors are caused by the absorption and scattering of light in water resulting in a blue-green hue and fog-like appearance. It is important to comprehend these challenges in order to develop techniques to overcome them and improve the quality of underwater video images..

[Audio] Moving on to slide number 8 our focus will be on the challenges that arise in underwater imaging. These obstacles include color distortion low visibility due to haze and turbidity light absorption blur and distortion noise non-uniform illumination limited ground truth data real-time processing depth variations and environmental variability. Moreover hardware limitations and data transmission can also have an impact on the quality of underwater imaging. As professionals in information technology electronics and intelligent communication systems it is crucial for us to understand and address these challenges in order to enhance the capabilities of underwater imaging. Let's now proceed to the next slide to further explore this topic..

[Audio] In our presentation we will discuss the importance of improving underwater images in the fields of information technology electronics and intelligent communication systems. The aquatic environment presents unique challenges for capturing images such as reduced visibility color distortion and noise. These challenges are caused by factors like water turbidity light absorption and scattering. Despite these difficulties it is essential to enhance underwater images for applications like marine biology underwater archaeology environmental monitoring and autonomous underwater vehicle navigation. Better image quality allows for more valuable information to be extracted and aids researchers and engineers in their studies. Effective image enhancement techniques like filtering and correction are necessary to achieve this goal. These techniques play a crucial role in revealing details enhancing contrast and mitigating the effects of underwater imaging. By addressing the challenges of the aquatic environment advancements in information technology electronics and intelligent communication systems can be made. In conclusion improving underwater images is vital for research and exploration in the underwater world. Therefore understanding the challenges and implementing effective techniques is crucial for progress in this field. Thank you for listening and we hope you keep these points in mind as we continue our presentation..

Proposed Methodology. Underwater Image Red Channel Green Channel Blue Channel Median Filtering Average Filtering Wiener Filtering Gaussian Filtering Wavelet Transform based Filtering Enhanced UWI Image Performance Evaluation.

[Audio] In this slide we will be discussing Algorithm 1: Median Filtering for U-W-I enhancement. This algorithm is used to enhance digital pictures and reduce noise in digital photographs. Unlike linear filters which use weighted averages the median filter replaces each pixel's value with the median value found in a specific neighborhood or window. This approach is particularly useful in minimizing impulsive noise such as salt-and-pepper noise which occurs when individual pixels have extreme values. The median filter is known for its exceptional ability to preserve edges and fine details in pictures making it well-suited for applications where it is crucial to maintain the integrity of object borders and complex features. This is due to its non-linear nature which makes it less susceptible to outliers and contributes to its robust performance in various imaging scenarios. However the window size selection in median filtering is a critical factor that greatly affects the balance between noise reduction and picture preservation. It is important to carefully consider the window size in order to achieve the desired results. In summary the median filter is a highly effective tool for enhancing digital pictures and reducing noise in digital photographs. Its ability to preserve edges and fine details along with its robust performance in different imaging situations makes it a valuable tool in the field of information technology electronics and intelligent communication systems..

[Audio] In this presentation we have discussed various techniques and methods in the field of image processing. Now let's focus on one of the most commonly used techniques average filtering. This algorithm also known as mean filtering is used to smooth or blur digital images. It works by replacing a pixel's value with the average value of the surrounding pixels within a designated window. This window typically a square with odd dimensions is moved across the entire image averaging the pixel values within it. The result is a reduction of high-frequency noise and a decrease in pixel intensity variations. This creates a smoother version of the original image without compromising important details. Despite its simplicity average filtering remains a valuable tool in image preprocessing and is often combined with more complex methods. Remember in the world of image processing sometimes simple techniques can make a significant impact. Thank you for listening and stay tuned for more interesting methods in our presentation..

[Audio] Welcome to slide number 13 of our presentation on the 2024 International Conference on Information Technology Electronics and Intelligent Communication Systems. Today we will be discussing Algorithm 3: Weiner Filtering for U-W-I enhancement. Wiener filtering is a commonly used method for improving image quality especially when dealing with images affected by additive noise. This technique aims to create a filter that reduces the mean square error between the original and filtered image. By implementing the Wiener filter we can minimize the impact of noise resulting in clearer and cleaner images. This is particularly useful in underwater scenes with turbidity scattering and light absorption which can greatly reduce image visibility. The filter effectively reveals hidden details making it a powerful tool for image enhancement in various scenarios particularly in underwater environments. Let's move on to our next slide..

[Audio] Welcome to our presentation on the 2024 International Conference on Information Technology Electronics and Intelligent Communication Systems. We're currently on slide 14 out of 24 and will be discussing Algorithm 4: Gaussian Filtering for U-W-I enhancement. This algorithm applies a Gaussian filter twice to each color channel at the beginning of the procedure. A Gaussian filter is a type of low pass filter that removes high-frequency components in the liver's U-W-I--. By using this filter we obtain a two-dimensional U-W-I without these components. To demonstrate this mathematically we have provided Equation 1 which describes the filtering process. The variables σ x y and z represent the standard deviation of the Gaussian distribution the position of pixels within the Gaussian window and the number of pixels in the image respectively. By adjusting these variables we can produce a U-W-I with smooth edges. Varying the window width is necessary to preserve the original integrity of the U-W-I--. This can be achieved by setting the variance parameter σ to one of three unique values depending on the desired effect. The table on this slide illustrates the different values for σ and their corresponding effects on the U-W-I--. This step is crucial in enhancing the overall quality and clarity of the U-W-I--. In conclusion we have discussed the use of a Gaussian filter in Algorithm 4 for U-W-I enhancement. This filter eliminates high-frequency components and produces a smooth U-W-I--. Thank you for your attention. I will now pass it on to my colleague for the next part of our presentation..

Algorithm 5: Wavelet Filtering for UWI enhancement.

[Audio] As we discuss the 2024 International Conference on Information Technology Electronics and Intelligent Communication Systems our focus now turns to slide number 16. This slide showcases the Image Enhancement Benchmark with a total of 950 real world underwater images 890 of which have corresponding reference images. This dataset is available to the public at the website github benchmark.html for those interested. On the slide you can see two types of images raw and reference. This benchmark plays a crucial role in evaluating and improving image enhancement techniques in the field of underwater imaging. Our next slide will further explore the significance of this benchmark and its impact on the industry..

Performance Evaluation.

Experimental Results. [image] Original Image Weiner Filtered Image Gaussian Filtered Image Average Filtered Image Median Filtered Image Wavelet Filtered Image.

[Audio] Welcome everyone to our presentation on the 2024 International Conference on Information Technology Electronics and Intelligent Communication Systems. We will be discussing slide number 19 out of 24 titled "Paper ID: PRESENTER name Affiliation". On this slide you will see a table displaying data on different methods and their corresponding M-S-E P-S-N-R AG P-C-Q-I and uciqe values. This data shows the comparison of Gaussian Median Average Wiener and Wavelet methods. Each method yields varying results highlighting the importance of selecting the appropriate method for image processing. Let's continue to the next slide..

[Audio] Today we will discuss the findings from our study on techniques used in Underwater Image Enhancement (U-W-I-E-). Specifically we will focus on the strengths and weaknesses of the Median Filter and the Average Filter. Beginning with the Median Filter we found that it effectively reduces impulse noise and outliers in underwater images while preserving edges and details. However it may not be as effective for other types of noise and may not be suitable for all underwater scenes. Moving on to the Average Filter we found it to be a simple and efficient technique for reducing high-frequency noise. However it may blur edges and details and is less effective against impulse noise. In conclusion it is important to consider the type of noise before choosing the appropriate filter. Thank you for your attention and we will continue discussing the remaining slides..

[Audio] Continuing our discussion on image denoising techniques we will now focus on two commonly used methods: the Wiener filter and the Gaussian filter. Each of these approaches has its own advantages and limitations which we will examine closely. The Wiener filter is an adaptive filter that minimizes mean square error making it effective in reducing noise while preserving important image details. However it relies on knowledge of noise statistics which can be difficult to acquire in real-world situations. Inaccurate noise estimation can also impact its performance. The Gaussian filter on the other hand is known for its smoothing capabilities while still retaining edges to some degree. It also allows for adjustment to control the amount of blurring in an image. However it may not be suitable for non-Gaussian noise and may have limited ability to enhance details in noisy images. It is crucial to carefully consider the strengths and limitations of these techniques before choosing one for a specific scenario. In conclusion we will discuss other image denoising techniques and their respective pros and cons. Thank you for your attention on slide number 21 and I look forward to our continued discussion on ICITEICS..

[Audio] Welcome to my presentation on the 2024 International Conference on Information Technology Electronics and Intelligent Communication Systems. We are currently on slide number 22. Today we will discuss the findings from a study on various uwie techniques specifically the Wavelet Filter technique. This technique has notable strengths such as its ability to perform multiresolution analysis and preserve image details while reducing noise. It is also effective for handling different scales of features in underwater images. However like any technique there are also weaknesses to consider. The use of Wavelet Filters is computationally intensive and may impact practical application. Selecting the appropriate wavelet and decomposition levels can also be challenging. Overall the Wavelet Filter technique shows potential for improving image quality in underwater environments. I look forward to discussing this further during our Q&A session..

[Audio] In this presentation we will discuss an important topic in information technology electronics and communication systems: underwater image enhancement. Advanced techniques and algorithms have been developed but there is still room for improvement. We will be looking at the latest research in this area including global and local equalization of histograms dual-image multi-scale fusion hyper-laplacian reflectance priors and minimal color loss and locally adaptive contrast enhancement. Notable papers include Bai and others's work on an underwater image enhancement benchmark dataset and beyond and Zhuang and others's use of hyper-laplacian reflectance priors. Other innovative approaches include twin adversarial contrastive learning fully guided information flow networks and the U-shape transformer all of which have shown promising results. We will also discuss benchmark datasets and objective metrics used for evaluation. Near the end of the presentation we will highlight Zhou and others's multi-view underwater image enhancement network which considers different perspectives and utilizes multi-view information for improved results. With technology constantly advancing we anticipate even more progress in this field. That concludes our presentation thank you for your attention. If you have any questions or comments please feel free to ask during the Q&A session..

THANK YOU.