scholarly journals Underwater Image Enhancement with a Deep Residual Framework

Author(s):  
Anuja Phapale ◽  
Puja Kasture ◽  
Keshav Katkar ◽  
Omkar Karale ◽  
Atal Deshmukh

This paper focuses on framework developed with the goal to enhance the quality of underwater images using machine learning models for the Underwater Image enhancement system. It also covers the various technologies and language used in the development process using Python programming language. The developed system provides two major functionality such as feature to provide input as image or video and returns enhanced image or video depending upon user input type with focus on more image quality, sharpness, colour correctness etc.

2021 ◽  
Vol 12 (5) ◽  
pp. 233-254
Author(s):  
D. Yu. Bulgakov ◽  

A method for solving resource-intensive tasks that actively use the CPU, when the computing resources of one server become insufficient, is proposed. The need to solve this class of problems arises when using various machine learning models in a production environment, as well as in scientific research. Cloud computing allows you to organize distributed task processing on virtual servers that are easy to create, maintain, and replicate. An approach based on the use of free software implemented in the Python programming language is justified and proposed. The resulting solution is considered from the point of view of the theory of queuing. The effect of the proposed approach in solving problems of face recognition and analysis of biomedical signals is described.


Author(s):  
Noé Sturm ◽  
Jiangming Sun ◽  
Yves Vandriessche ◽  
Andreas Mayr ◽  
Günter Klambauer ◽  
...  

<div>This article describes an application of high-throughput fingerprints (HTSFP) built upon industrial data accumulated over the years. </div><div>The fingerprint was used to build machine learning models (multi-task deep learning + SVM) for compound activity predictions towards a panel of 131 targets. </div><div>Quality of the predictions and the scaffold hopping potential of the HTSFP were systematically compared to traditional structural descriptors ECFP. </div><div><br></div>


Author(s):  
Jože M. Rožanec ◽  
Elena Trajkova ◽  
Jinzhi Lu ◽  
Nikolaos Sarantinoudis ◽  
Georgios Arampatzis ◽  
...  

Refineries execute a series of interlinked processes, where the product of one unit serves as the input to another process. Potential failures within these processes affect the quality of the end products, operational efficiency, and revenue of the entire refinery. In this context, implementation of a real-time cognitive module, referring to predictive machine learning models, enables to provide equipment state monitoring services and to generate decision-making for equipment operations. In this paper, we propose two machine learning models: 1) to forecast the amount of pentane (C5) content in the final product mixture; 2) to identify if C5 content exceeds the specification thresholds for the final product quality. We validate our approach by using a use case from a real-world refinery. In addition, we develop a visualization to assess which features are considered most important during feature selection, and later by the machine learning models. Finally, we provide insights on the sensor values in the dataset, which help to identify the operational conditions for using such machine learning models.


2021 ◽  
Vol 11 (24) ◽  
pp. 11790
Author(s):  
Jože Martin Rožanec ◽  
Elena Trajkova ◽  
Jinzhi Lu ◽  
Nikolaos Sarantinoudis ◽  
George Arampatzis ◽  
...  

Refineries execute a series of interlinked processes, where the product of one unit serves as the input to another process. Potential failures within these processes affect the quality of the end products, operational efficiency, and revenue of the entire refinery. In this context, implementation of a real-time cognitive module, referring to predictive machine learning models, enables the provision of equipment state monitoring services and the generation of decision-making for equipment operations. In this paper, we propose two machine learning models: (1) to forecast the amount of pentane (C5) content in the final product mixture; (2) to identify if C5 content exceeds the specification thresholds for the final product quality. We validate our approach using a use case from a real-world refinery. In addition, we develop a visualization to assess which features are considered most important during feature selection, and later by the machine learning models. Finally, we provide insights on the sensor values in the dataset, which help to identify the operational conditions for using such machine learning models.


2018 ◽  
Vol 12 (1) ◽  
pp. 810-823 ◽  
Author(s):  
Mohamad Javad Alizadeh ◽  
Mohamad Reza Kavianpour ◽  
Malihe Danesh ◽  
Jason Adolf ◽  
Shahabbodin Shamshirband ◽  
...  

2014 ◽  
Vol 926-930 ◽  
pp. 1704-1707
Author(s):  
Qiu Yun Wang

According to the image formation model and the nature of underwater images, we find that the effect of the haze and the color distortion seriously pollute the underwater image data, lowing the quality of the underwater images in the visibility and the quality of the data. Hence, aiming to reduce the noise and the haze effect existing in the underwater image and compensate the color distortion, the dark channel prior model is used to enhance the underwater image. We compare the dark channel prior model based image enhancement method to the contrast stretching based method for image enhancement. The experimental results proved that the dark channel prior model has good ability for processing the underwater images. The super performance of the proposed method is demonstrated as well.


Author(s):  
Carmel Kent ◽  
Muhammad Ali Chaudhry ◽  
Mutlu Cukurova ◽  
Ibrahim Bashir ◽  
Hannah Pickard ◽  
...  

2021 ◽  
Author(s):  
Michael Tarasiou

This paper presents DeepSatData a pipeline for automatically generating satellite imagery datasets for training machine learning models. We also discuss design considerations with emphasis on dense classification tasks, e.g. semantic segmentation. The implementation presented makes use of freely available Sentinel-2 data which allows the generation of large scale datasets required for training deep neural networks (DNN). We discuss issues faced from the point of view of DNN training and evaluation such as checking the quality of ground truth data and comment on the scalability of the approach.


2017 ◽  
Vol 22 (3) ◽  
pp. 31-38
Author(s):  
Ritu Singh ◽  
Mantosh Biswas

Abstract Scattering and absorption of light in water leads to degradation of images captured under the water. This degradation includes diminished colors, low brightness and undistinguishable objects in the image. To improve the quality of such degraded images, we have proposed fusion based underwater image enhancement technique that focuses on improving of the contrast and color of underwater images using contrast stretching and Auto White Balance. Our proposed method is very simple and straightforward that contributes greatly in uplifting the visibility of underwater images.


Sign in / Sign up

Export Citation Format

Share Document