Large Image Databases Research Articles

A METHOD FOR SYNTHESIZING THERMAL IMAGES USING GAN MULTI-LAYERED APPROACH

Abstract. The active development of neural network technologies and optoelectronic systems has led to the introduction of computer vision technologies in various fields of science and technology. Deep learning made it possible to solve complex problems that a person had not been able to solve before. The use of multi-spectral optical systems has significantly expanded the field of application of video systems. Tasks such as image recognition, object re-identification, video surveillance require high accuracy, speed and reliability. These qualities are provided by algorithms based on deep convolutional neural networks. However, they require to have large databases of multi-spectral images of various objects to achieve state-of-the-art results. While large and various databases of color images of different objects are widely available in public domain, then similar databases of thermal images are either not available, or they represent a small number of types of objects. The quality of three-dimensional modeling for the thermal imaging spectral range remains at an insufficient level for solving a number of important tasks, which require high precision and reliability. The realistic synthesis of thermal images is especially important due to the complexity and high cost of obtaining real data. This paper is focused on the development of a method for synthesizing thermal imaging images based on generative adversarial neural networks. We developed an algorithm for a multi-spectral image-to-image translation. We have changed to the original GAN architecture and converted the loss function. We presented a new learning approach. For this, we prepared a special training dataset including about 2000 image tensors. The evaluation of the results obtained showed that the proposed method can be used to expand the available databases of thermal images.

An analysis of content-based image retrieval

Nowadays, working on digital images is gaining much popularity in multimedia systems, due to the rapid increase in the utilization of large image databases. Thus, the Content-Based Image Retrieval (CBIR) method has become the most valuable method for these databases. This study mainly focuses on content-based image retrieval; which uses image features like color, shape, texture, etc. by searching the user query image from a large image database based on user request. CBIR is the most widely used technique as its searching capability is faster than the other traditional methods, and it works well in retrieving images automatically. It is also a big alternative approach to traditional methods. The CBIR techniques are used in many applications like surveillance detection, crime avoidance, fingerprint identification, E-library, medical, historical monument and biodiversity information systems, and many more. A total of 38 CBIR articles were comparatively analyzed.

A deep learning approach for the automatic recognition of prosthetic mitral valve in echocardiographic images

The first step in the automatic evaluation of the cardiac prosthetic valve is the recognition of such valves in echocardiographic images. This research surveyed whether a deep convolutional neural network (DCNN) could improve the recognition of prosthetic mitral valve in conventional 2D echocardiographic images. An efficient intervention to decrease the misreading rate of the prosthetic mitral valve is required for non-expert cardiologists. This intervention could serve as a section of a fully-automated analysis chain, alleviate the cardiologist's workload, and improve precision and time management, especially in an emergent situation. Additionally, it might be suitable for pre-labeling large databases of unclassified images. We, therefore, introduce a large publicly-available annotated dataset for the purpose of prosthetic mitral valve recognition. We utilized 2044 comprehensive non-stress transthoracic echocardiographic studies. Totally, 1597 patients had natural mitral valves and 447 patients had prosthetic valves. Each case contained 1 cycle of echocardiographic images from the apical 4-chamber (A4C) and the parasternal long-axis (PLA) views. Thirteen versions of the state-of-the-art models were independently trained, and the ensemble predictions were performed using those versions. For the recognition of prosthetic mitral valves from natural mitral valves, the area under the receiver-operating characteristic curve (AUC) made by the deep learning algorithm was similar to that made by cardiologists (0.99). In this research, EfficientNetB3 architecture in the A4C view and the EfficientNetB4 architecture in the PLA view were the best models among the other pre-trained DCNN models.

Blind 360-degree image quality assessment via saliency-guided convolution neural network

With the rapid development of virtual reality (VR) technologies, quality assessment of 360-degree images has become increasingly urgent. Unlike the traditional 2D images, the distortion is not evenly distributed in the 360-degree images, e.g., the projection distortion in the polar region of the Equirectangular projection (ERP) is more serious than that in other regions. Thus, the traditional 2D quality model cannot be directly applied to 360-degree images. In this paper, we propose a saliency-guided CNN model for blind 360-degree image quality assessment (SG360BIQA), which is mainly composed of a saliency prediction network (SP-Net) and a feature extraction network (F-Net). By training the whole network with the two sub-networks together, more discriminant features can be extracted and the mapping from feature representations to quality scores can be established more accurately. Besides, due to the lack of sufficient large 360-degree images database, we use a pre-trained network model instead of the initial random parameter model to overcome the limitation. Experimental results on two public 360-IQA databases demonstrate that our proposed model outperforms state-of-the-art full-reference and no-reference IQA metrics in terms of generalization ability and evaluation accuracy.

Content-Based Image Retrieval Using Discrete Wavelet Transform and Convolutional Neural Network

Due to the development of internet and digital media techniques the size of digital image collection is increasing rapidly. Various techniques for storing, searching and retrieving images become essential for large image archives. This leads to an existence of Image Retrieval System. Content-Based Image Retrieval (CBIR) is a process of retrieving the expected images from large image databases based on the features of the query image and this is performed by various feature extraction techniques. In this work, two different methods of CBIR have been proposed. The first method is CBIR using Discrete Wavelet Transform (DWT), which extracts the texture feature of an image and histogram extracts the color feature of an image. And also the results will be ranked according to the chi-square distance calculation. The second method is CBIR using Convolutional Neural Network (CNN). In CNN method the features are extracted using VGG16 pre-trained model and canny edge detection is used for detecting the edges of an image. Finally, the performances of DWT and CNN based methods are analyzed and compared.

Discourse and Identity in the Digital World: Mansoura University Website as a Model

Digital Discourse is concerned with the manipulation of multi-modal and multi-semiotic resources which are investigated to characterize identities and ideologies in a digital world that is said to be a part of a whole society, as claimed by Gee (2005). A digital text is a type of human communication that can be described as multimodal by incorporating writing, images, sounds and other semiotic systems. In the last three decades, people have become more attracted to say so many things about themselves and their activities through digital media. However, attempts to give a comprehensive image of a college or university are limited. The present study aims to signify how a digital text can reflect ideologies and realities about the social and academic life of Mansoura University, located in Mansoura city, the capital of Dakahlia Governorate on the east bank of the Nile in Egypt. The data employed as the basis of the analysis in this study is extracted from large database of announcements, texts and images displayed on Mansoura University website in 2020 (following the widespread transmission of Covid-19). The study proposes a multidisciplinary analysis of Mansoura University website, following Darvin (2016) and Petroni (2019). Mansoura University website is said to be an attempt by a collective institution to establish an image of itself on the internet. The digitized discourse is designed to provide a collective image that replicates the identity of Mansoura University and highlights its position (inter)nationally. Dealing with the information displayed on the target website as a type of commodity that possesses multiple traits and, at the same time, reflects issues of identity and reputation building, the study concludes that the analysis of the digital discourse has assigned Mansoura University a type of identity that can be described as a Commodified Identity.

Differential Deep Convolutional Neural Network Model for Brain Tumor Classification.

The classification of brain tumors is a difficult task in the field of medical image analysis. Improving algorithms and machine learning technology helps radiologists to easily diagnose the tumor without surgical intervention. In recent years, deep learning techniques have made excellent progress in the field of medical image processing and analysis. However, there are many difficulties in classifying brain tumors using magnetic resonance imaging; first, the difficulty of brain structure and the intertwining of tissues in it; and secondly, the difficulty of classifying brain tumors due to the high density nature of the brain. We propose a differential deep convolutional neural network model (differential deep-CNN) to classify different types of brain tumor, including abnormal and normal magnetic resonance (MR) images. Using differential operators in the differential deep-CNN architecture, we derived the additional differential feature maps in the original CNN feature maps. The derivation process led to an improvement in the performance of the proposed approach in accordance with the results of the evaluation parameters used. The advantage of the differential deep-CNN model is an analysis of a pixel directional pattern of images using contrast calculations and its high ability to classify a large database of images with high accuracy and without technical problems. Therefore, the proposed approach gives an excellent overall performance. To test and train the performance of this model, we used a dataset consisting of 25,000 brain magnetic resonance imaging (MRI) images, which includes abnormal and normal images. The experimental results showed that the proposed model achieved an accuracy of 99.25%. This study demonstrates that the proposed differential deep-CNN model can be used to facilitate the automatic classification of brain tumors.

Detection of Image Forgery Using Information Standard Method With SVM

Depending on various features, image forgery is now used to enhance and restore images to make them more significant, whereas image forgery is performed by vulnerable people to generate fake truths by tampering images. From large (image) databases, a specified number of images forgery, new mechanism will have implemented to detect systems which is mainly based on two procedures. The first procedure relies on extract features of image through filters (LBP, Gabor, PCA and DCT) to find out its efficiency evaluated using some Statistics Measurements(Mean, Variance) and the second procedure relies on several quality metrics in optimization processes which (PSNR, MSE, AIC) have been calculated to which results in better quality compared to other image by combining selected filters with SVM, and, through experimental test, it was found that this proposed method of detect forgery technique is powerful than the classical detect system. The proposed methodology gives the better accuracy up to 98.5%, to detect and extract all type of forgery, with image enhancement capabilities.

Content Based Image Retrieval Using Deep Learning Convolutional Neural Network

Content-based image retrieval (CBIR) is a widely used method for image retrieval from large and unlabeled image collections. However, users are not satisfied with the traditional methods of retrieving information. Moreover the abundance of online networks for production and distribution, as well as the quantity of images accessible to consumers, continues to expand. Therefore, in many areas, permanent as well as widespread digital image processing takes place. Therefore, the rapid access to these large image databases as well as the extraction of identical images from this large set of images from a given image (Query) pose significant challenges as well as involves efficient techniques. A CBIR system’s efficiency depends fundamentally on the calculation of feature representation as well as similarity. For this purpose, they present a basic but powerful deep learning system focused on Convolutional Neural Networks (CNN) and composed of feature extraction and classification for fast image retrieval. We get some promising findings from many detailed observational studies for a number of CBIR tasks using image database, which reveals some valuable lessons for improving the efficiency of CBIR. CBIR systems allow another image dataset to locate related images to such a query image. The search per picture function of Google search has to be the most popular CBIR method.

An automated framework for image classification and segmentation of fetal ultrasound images for gestational age estimation.

Accurate assessment of fetal gestational age (GA) is critical to the clinical management of pregnancy. Industrialized countries rely upon obstetric ultrasound (US) to make this estimate. In low- and middle- income countries, automatic measurement of fetal structures using a low-cost obstetric US may assist in establishing GA without the need for skilled sonographers. In this report, we leverage a large database of obstetric US images acquired, stored and annotated by expert sonographers to train algorithms to classify, segment, and measure several fetal structures: biparietal diameter (BPD), head circumference (HC), crown rump length (CRL), abdominal circumference (AC), and femur length (FL). We present a technique for generating raw images suitable for model training by removing caliper and text annotation and describe a fully automated pipeline for image classification, segmentation, and structure measurement to estimate the GA. The resulting framework achieves an average accuracy of 93% in classification tasks, a mean Intersection over Union accuracy of 0.91 during segmentation tasks, and a mean measurement error of 1.89 centimeters, finally leading to a 1.4 day mean average error in the predicted GA compared to expert sonographer GA estimate using the Hadlock equation.

Human Symmetry Uncertainty Detected by a Self-Organizing Neural Network Map

Symmetry in biological and physical systems is a product of self-organization driven by evolutionary processes, or mechanical systems under constraints. Symmetry-based feature extraction or representation by neural networks may unravel the most informative contents in large image databases. Despite significant achievements of artificial intelligence in recognition and classification of regular patterns, the problem of uncertainty remains a major challenge in ambiguous data. In this study, we present an artificial neural network that detects symmetry uncertainty states in human observers. To this end, we exploit a neural network metric in the output of a biologically inspired Self-Organizing Map Quantization Error (SOM-QE). Shape pairs with perfect geometry mirror symmetry but a non-homogenous appearance, caused by local variations in hue, saturation, or lightness within and/or across the shapes in a given pair produce, as shown here, a longer choice response time (RT) for “yes” responses relative to symmetry. These data are consistently mirrored by the variations in the SOM-QE from unsupervised neural network analysis of the same stimulus images. The neural network metric is thus capable of detecting and scaling human symmetry uncertainty in response to patterns. Such capacity is tightly linked to the metric’s proven selectivity to local contrast and color variations in large and highly complex image data.

Image Retrieval System-An Integrated Approach

With the development of multimedia technology, the usage of large image database becomes possible and is rapidly increasing. These can be used for the purpose of retrieving files, which works on the mechanism of image search. Different databases are available on different websites like Instagram, Facebook, Twitter, Flickr, and Picasa. This paper shows the advantage of content-based image retrieval system, as well as its key technologies. Comparing to the shortcoming, only certain feature are used in the traditional system. This paper presents a review on different techniques of image retrieval techniques which are based on color, texture and shape of images. It also focuses on the feature extraction and representation, several commonly used algorithms and different methods used for matching of images.

A Novel Filtered Segmentation-Based Bayesian Deep Neural Network Framework on Large Diabetic Retinopathy Databases

Image thresholding-based segmentation models play a vital role in the detection of Diabetic retinopathy (DR) on large databases. Most of the conventional segmentation-based classification models are independent of over segmented regions and outliers. Also, these models have less true positive rate and high error rate on different DR feature sets. In order to overcome these problems, a novel filtered based segmentation framework is designed and implemented on the large DR feature space. In this work, a novel image filtering approach, optimal image segmentation approach and hybrid Bayesian deep learning framework are developed on the large DR image databases. Experimental results proved that the proposed filtered segmentation-based Bayesian deep neural network has better accuracy and runtime than the conventional models on different DR variation databases.

Dermatological Disease Classification utilizing Image Processing and Neural Networks

Skin diseases are a frequent problem among all age groups. Application of Machine Learning (ML) is exceedingly suitable for skin diseases identification as it has large clinical image database that can be used to train models and interpret diagnosis for better patient outcomes. Researchers have used various image processing techniques and classification methods. Color and Texture based features are most commonly used for image analysis and Convolutional Neural Network (CNN) has become current standard practice in classifying skin disorders. This paper presents a thorough survey of image processing techniques and classifiers for skin diseases detection.

QuoVidi: An open-source web application for the organization of large-scale biological treasure hunts.

Learning biology, and in particular systematics, requires learning a substantial amount of specific vocabulary, both for botanical and zoological studies. While crucial, the precise identification of structures serving as evolutionary traits and systematic criteria is not per se a highly motivating task for students. Teaching this in a traditional teaching setting is quite challenging especially with a large crowd of students to be kept engaged. This is even more difficult if, as during the COVID‐19 crisis, students are not allowed to access laboratories for hands‐on observation on fresh specimens and sometimes restricted to short‐range movements outside their home. Here, we present QuoVidi, a new open‐source web platform for the organization of large‐scale treasure hunts. The platform works as follows: students, organized in teams, receive a list of quests that contain morphologic, ecologic, or systematic terms. They have to first understand the meaning of the quests, then go and find them in the environment. Once they find the organism corresponding to a quest, they upload a geotagged picture of their finding and submit this on the platform. The correctness of each submission is evaluated by the staff. During the COVID‐19 lockdown, previously validated pictures were also submitted for evaluation to students that were locked in low‐biodiversity areas. From a research perspective, the system enables the creation of large image databases by the students, similar to citizen science projects. Beside the enhanced motivation of students to learn the vocabulary and perform observations on self‐found specimens, this system allows instructors to remotely follow and assess the work performed by large numbers of students. The interface is freely available, open‐source and customizable. Unlike existing naturalist platforms, allows the educators to fully customize the quests of interest. This enables the creation of multiple teaching scenarios, without being bound to a fixed scope. QuoVidi can be used in other disciplines with adapted quests and we expect it to be of interest in many classroom settings.

New Image Classification Framework for Improving Image Retrieval Based on Relevance Feedback (Dept.E)

Content-based image retrieval (CBIR) is a promising technology to assist image finding, CBIR retrieves images by visual features inherent in images. Relevance feedback allows the user to reflect his preference to the system, then the system can reformulate the query according to the positive and/or negative examples responded by the user. This paper presents two efficient frameworks for image classification through the analysis of the visual features (such as color, shape, texture) of an example image. The first framework is based on comparing the norm and the direction of vectors ia multi-dimensional space for both the target and query image vectors, which allows classification according to their probabilities oli existence. The second framework depends on acquiring classification knowledge from a large empirical image database in a specific domain and utilizes that knowledge for image classification. The initial classification process is used as a training phase to feed the system with a classification tree for images in the retrieval domain. This tree is the best reduction of dimensionality that would result from all possible combinations of feature divisions.

Exploring the Impact of Similarity Model to Identify the Most Similar Image from a Large Image Database

Identifying the most similar image from a given large image database is becoming more and more popular, and many companies have brought out related products. However, there are some incomplete functions and shortcoming of previous work, the database was disorganized and does not have any regulation, and the similarity method used in the image searching system can be improved. In this study, we have implemented an improved similar image searching system that can filter the dataset based on the requirement of users. Besides, an experiment about testing different similarity models has been done. The result is an improved similarity model which can be subsequently applied to the searching system.

Computational approach for content‐based image retrieval of K‐similar images from brain MR image database

AbstractContent‐based medical image retrieval (CBMIR) is a mechanism to handle a huge quantity of image data generated in various medical imaging modalities. In recent years, due to the evolution of computer vision and digital imaging modalities, a large number of medical images are generated. Consequently, the task of retrieving medical images from a large image database becomes more tedious due to variation in the size and shape of the images. Hence, it is necessary to design an appropriate system for medical image retrieval. In this paper a methodology for CBMIR using features of an image such as colour, shape, and texture is proposed to represent and retrieve the images from a large database that are relevant to a given query image. This methodology is evaluated for the application of retrieving the brain MRI images of different planes (coronal, sagittal, and transverse) from a dataset of normal and demented subjects. The features are determined in terms of Grey level co‐occurrence based Haralik's features and histogram based cumulative distribution function (CDF). The image retrieval mechanism is designed using the K‐Nearest Neighbour algorithm by finding the minimum distance between query and database images. The performance parameters such as precision and recall are calculated. The average accuracy of 95.5% are obtained. The results provided ensures the capability to use it as assistive framework for radiologists in radiology image retrieval and classification.

Unsupervised quality control of segmentations based on a smoothness and intensity probabilistic model.

Monitoring the quality of image segmentation is key to many clinical applications. This quality assessment can be carried out by a human expert when the number of cases is limited. However, it becomes onerous when dealing with large image databases, so partial automation of this process is preferable. Previous works have proposed both supervised and unsupervised methods for the automated control of image segmentations. The former assume the availability of a subset of trusted segmented images on which supervised learning is performed, while the latter does not. In this paper, we introduce a novel unsupervised approach for quality assessment of segmented images based on a generic probabilistic model. Quality estimates are produced by comparing each segmentation with the output of a probabilistic segmentation model that relies on intensity and smoothness assumptions. Ranking cases with respect to these two assumptions allows the most challenging cases in a dataset to be detected. Furthermore, unlike prior work, our approach enables possible segmentation errors to be localized within an image. The proposed generic probabilistic segmentation method combines intensity mixture distributions with spatial regularization prior models whose parameters are estimated with variational Bayesian techniques. We introduce a novel smoothness prior based on the penalization of the derivatives of label maps which allows an automatic estimation of its hyperparameter in a fully data-driven way. Extensive evaluation of quality control on medical and COCO datasets is conducted, showing the ability to isolate atypical segmentations automatically and to predict, in some cases, the performance of segmentation algorithms.

A narrative review of digital pathology and artificial intelligence: focusing on lung cancer.

The emergence of whole slide imaging technology allows for pathology diagnosis on a computer screen. The applications of digital pathology are expanding, from supporting remote institutes suffering from a shortage of pathologists to routine use in daily diagnosis including that of lung cancer. Through practice and research large archival databases of digital pathology images have been developed that will facilitate the development of artificial intelligence (AI) methods for image analysis. Currently, several AI applications have been reported in the field of lung cancer; these include the segmentation of carcinoma foci, detection of lymph node metastasis, counting of tumor cells, and prediction of gene mutations. Although the integration of AI algorithms into clinical practice remains a significant challenge, we have implemented tumor cell count for genetic analysis, a helpful application for routine use. Our experience suggests that pathologists often overestimate the contents of tumor cells, and the use of AI-based analysis increases the accuracy and makes the tasks less tedious. However, there are several difficulties encountered in the practical use of AI in clinical diagnosis. These include the lack of sufficient annotated data for the development and validation of AI systems, the explainability of black box AI models, such as those based on deep learning that offer the most promising performance, and the difficulty in defining the ground truth data for training and validation owing to inherent ambiguity in most applications. All of these together present significant challenges in the development and clinical translation of AI methods in the practice of pathology. Additional research on these problems will help in resolving the barriers to the clinical use of AI. Helping pathologists in developing knowledge of the working and limitations of AI will benefit the use of AI in both diagnostics and research.