Automatic Image Recognition Research Articles

Medical image recognition based on VGGNet19

Gastrointestinal diseases are one of the common clinical diseases, which often require medical imaging for diagnosis and treatment. Recently, the development of deep learning technology has promoted the development of medical image recognition, which provides new ideas and methods for the automatic recognition and analysis of medical images. VGGNet19 is a convolutional neural network model that has attracted much attention because of its simple structure, easy training and better recognition effect. For this reason, this paper proposes an improved VGGNet19 model for medical image recognition of gastrointestinal diseases. Specifically, the project adds an additional fully connected layer and Dropout layer on top of the built VGGNet19 to achieve the recognition of medical images of stomach diseases. Extensive experiments on standard medical stomach images show that the proposed method improves the recognition performance to a certain extent.

An Interpretable Three-Dimensional Artificial Intelligence Model for Computer-Aided Diagnosis of Lung Nodules in Computed Tomography Images.

Lung cancer is typically classified into small-cell carcinoma and non-small-cell carcinoma. Non-small-cell carcinoma accounts for approximately 85% of all lung cancers. Low-dose chest computed tomography (CT) can quickly and non-invasively diagnose lung cancer. In the era of deep learning, an artificial intelligence (AI) computer-aided diagnosis system can be developed for the automatic recognition of CT images of patients, creating a new form of intelligent medical service. For many years, lung cancer has been the leading cause of cancer-related deaths in Taiwan, with smoking and air pollution increasing the likelihood of developing the disease. The incidence of lung adenocarcinoma in never-smoking women has also increased significantly in recent years, resulting in an important public health problem. Early detection of lung cancer and prompt treatment can help reduce the mortality rate of patients with lung cancer. In this study, an improved 3D interpretable hierarchical semantic convolutional neural network named HSNet was developed and validated for the automatic diagnosis of lung cancer based on a collection of lung nodule images. The interpretable AI model proposed in this study, with different training strategies and adjustment of model parameters, such as cyclic learning rate and random weight averaging, demonstrated better diagnostic performance than the previous literature, with results of a four-fold cross-validation procedure showing calcification: 0.9873 ± 0.006, margin: 0.9207 ± 0.009, subtlety: 0.9026 ± 0.014, texture: 0.9685 ± 0.006, sphericity: 0.8652 ± 0.021, and malignancy: 0.9685 ± 0.006.

Automatic Face Recognition of Target Images Based on Deep Learning Algorithms

The current research results of face recognition have poor recognition accuracy, time-consuming and high information loss rate. A method of automatic face recognition based on depth learning algorithm is proposed. The three primary RGB images are transformed into color saturated HSV images, and the dynamic range is extended by changing the non-linearity according to the conversion results. Based on the extended situation, the membership function is used to map the target image to the blurred plane, and the contrast enhancement of the image is completed. The enhanced image is substituted into the image segmentation, and the image eigenvalues are extracted preliminarily. The seed of image segmentation is generated by the eigenvalues. The growing seed regions are merged by the color distance and the texture distance to achieve the target image segmentation. Based on image segmentation, LBP operator is used to extract the local texture features of face twice, and then a deep convolution network model is constructed. The shared weights of convolution network model and pooling down-sampling technique are used to reduce the complexity of the model. The top layer of the model forms a feature classification surface of face image, fuses the constraint conditions, and obtains the trained face. The deep convolution network model is used to extract features from face images and complete face recognition. The experimental results show that the method has good accuracy, high efficiency and low information loss rate.

Retracted: Digital Image Automatic Recognition and Tracking Based on Electronic Derotation Algorithm

Retracted: Digital Image Automatic Recognition and Tracking Based on Electronic Derotation Algorithm

Monitoring Inland Water Quantity Variations: A Comprehensive Analysis of Multi-Source Satellite Observation Technology Applications

The advancement of multi-source Earth observation technology has led to a substantial body of literature on inland water monitoring. This has resulted in the emergence of a distinct interdisciplinary field encompassing the application of multi-source Earth observation techniques in inland water monitoring. Despite this growth, few systematic reviews of this field exist. Therefore, in this paper, we offer a comprehensive analysis based on 30,212 publications spanning the years 1990 to 2022, providing valuable insights. We collected and analyzed fundamental information such as publication year, country, affiliation, journal, and author details. Through co-occurrence analysis, we identified country and author partnerships, while co-citation analysis revealed the influence of journals, authors, and documents. We employed keywords to explore the evolution of hydrological phenomena and study areas, using burst analysis to predict trends and frontiers. We discovered exponential growth in this field with a closer integration of hydrological phenomena and Earth observation techniques. The research focus has shifted from large glaciers to encompass large river basins and the Tibetan Plateau. Long-term research attention has been dedicated to optical properties, sea level, and satellite gravity. The adoption of automatic image recognition and processing, enabled by deep learning and artificial intelligence, has opened new interdisciplinary avenues. The results of the study emphasize the significance of long-term, stable, and accurate global observation and monitoring of inland water, particularly in the context of cloud computing and big data.

Experimental Research into the Uniaxial Compressive Strength of Low-Density Reef Limestone Based on Image Recognition.

Low-density reef limestone is widely distributed in tropical oceans; exploring its mechanical properties is of significance to practices in marine foundation engineering. In this research, laboratory experiments on low-density reef limestones with two different types of porous structures were conducted using image recognition methods to study the special mechanical properties of low-reef limestone. S¯ was defined as the parameter quantifying the pore geometry, and the calculation method of S¯ was optimized based on image recognition data. Finally, the influencing factors of uniaxial compressive strength (UCS) of low-density reef limestone were analyzed, and a modified formula considering pore structure was proposed. The results indicate the following: Image recognition methods were used to determine feasibility and convenience of capturing 2D pore geometric information of specimens. The optimization method of S¯ is conducive to improving automatic image recognition accuracy. Low-density reef limestones with different porous structures show small difference in porosity and density, while they exhibit large differences in pore sizes and UCS. The UCS of low-density reef limestone is found to be jointly influenced by pore structure and density (it increases with the decrease in parameter S¯ and increase in density). The results may provide help for those investigating the mechanical properties of reef limestone and practices in marine foundation engineering.

Transformer-based automated segmentation of recycling materials for semantic understanding in construction

Construction sites are incorporating cameras to gather imagery data for project management. While transformer-based deep models show promise in recognizing construction objects and understanding the environment, their use in construction images is largely unexplored. This paper presents a systematic evaluation of three state-of-the-art transformer-based models for automatic segmentation and recognition of construction images. Further, a two-stage model ensembling strategy based on model averaging and probability weighting is introduced and implemented for performance improvement. A dataset containing five classes of recycling materials on construction sites is created as a benchmark to compare their performance. The comparison results indicate the ensemble model could achieve encouraging results with a mIoU of 82.36% and mPA of 90.30%, which demonstrate superior segmentation performance on construction images.

Adversarial Robustness Enhancement of UAV-Oriented Automatic Image Recognition Based on Deep Ensemble Models

Deep neural networks (DNNs) have been widely utilized in automatic visual navigation and recognition on modern unmanned aerial vehicles (UAVs), achieving state-of-the-art performances. However, DNN-based visual recognition systems on UAVs show serious vulnerability to adversarial camouflage patterns on targets and well-designed imperceptible perturbations in real-time images, which poses a threat to safety-related applications. Considering a scenario in which a UAV is suffering from adversarial attack, in this paper, we investigate and construct two ensemble approaches with CNN and transformer for both proactive (i.e., generate robust models) and reactive (i.e., adversarial detection) adversarial defense. They are expected to be secure under attack and adapt to the resource-limited environment on UAVs. Specifically, the probability distributions of output layers from base DNN models in the ensemble are combined in the proactive defense, which mainly exploits the weak adversarial transferability between the CNN and transformer. For the reactive defense, we integrate the scoring functions of several adversarial detectors with the hidden features and average the output confidence scores from ResNets and ViTs as a second integration. To verify their effectiveness in the recognition task of remote sensing images, we conduct experiments on both optical and synthetic aperture radar (SAR) datasets. We find that the ensemble model in proactive defense performs as well as three popular counterparts, and both of the ensemble approaches can achieve much more satisfactory results than a single base model/detector, which effectively alleviates adversarial vulnerability without extra re-training. In addition, we establish a one-stop platform for conveniently evaluating adversarial robustness and performing defense on recognition models called AREP-RSIs, which is beneficial for the future research of the remote sensing field.

The image recognition of urban greening tree species based on deep learning and CAMP-MKNet model

The information of urban tree species resources is of vital significance to the planning and design of urban green spaces. Tree organs, such as the bark are used as the primary features of identifying tree species. However, traditional tree identification methods need to consume a lot of manpower and time costs. In addition, the application of machine image recognition technology to tree species recognition still has problems such as heavy data preprocessing workload, small number of tree species images, uneven distribution of categories, and low recognition accuracy. In order to promote the intelligent management of urban forestry and solve the above problems, it is necessary to establish an automatic image recognition model for urban greening tree species. We captured bark images of 21 urban afforestation tree species in their natural environment and constructed a dataset that was divided into a train set, validation set, and test set in the ratio of 7:1:2. Combining Channel Attention Module (CAM) with algorithms such as Spatial Pyramid Pooling (SPP) and Mixed Depthwise Dilated Convolutional Kernels. The core algorithm is Mixed Convolution Kernel (MK), and a CAMP-MKNet Convolutional Neural Network (CNN) is constructed as a bark image classification model for urban greening tree species. The overall accuracy of the generic models ranged from 41.06% to 82.03%, whereas the overall accuracy of the experimental CAMP-MKNet model was 84.25%, with lower prediction cost. Our study shows that the CAMP-MKNet CNN model with better prediction performance and computational cost and can provide crucial insights and technical support for developing automated urban tree species image recognition systems.

Pre-Trained Deep Neural Network-Based Features Selection Supported Machine Learning for Rice Leaf Disease Classification

Rice is a staple food for roughly half of the world’s population. Some farmers prefer rice cultivation to other crops because rice can thrive in a wide range of environments. Several studies have found that about 70% of India’s population relies on agriculture in some way and that agribusiness accounts for about 17% of India’s GDP. In India, rice is one of the most important crops, but it is vulnerable to a number of diseases throughout the growing process. Farmers’ manual identification of these diseases is highly inaccurate due to their lack of medical expertise. Recent advances in deep learning models show that automatic image recognition systems can be extremely useful in such situations. In this paper, we propose a suitable and effective system for predicting diseases in rice leaves using a number of different deep learning techniques. Images of rice leaf diseases were gathered and processed to fulfil the algorithmic requirements. Initially, features were extracted by using 32 pre-trained models, and then we classified the images of rice leaf diseases such as bacterial blight, blast, and brown spot with numerous machine learning and ensemble learning classifiers and compared the results. The proposed procedure works better than other methods that are currently used. It achieves 90–91% identification accuracy and other performance parameters such as precision, Recall Rate, F1-score, Matthews Coefficient, and Kappa Statistics on a normal data set. Even after the segmentation process, the value reaches 93–94% for model EfficientNetV2B3 with ET and HGB classifiers. The proposed model efficiently recognises rice leaf diseases with an accuracy of 94%. The experimental results show that the proposed procedure is valid and effective for identifying rice diseases.

Robust Semisupervised Federated Learning for Images Automatic Recognition in Internet of Drones

Air access networks have been recognized as a significant driver of various Internet of Things (IoT) services and applications. In particular, the aerial computing network infrastructure centered on the Internet of Drones has set off a new revolution in automatic image recognition. This emerging technology relies on sharing ground-truth-labeled data between unmanned aerial vehicle (UAV) swarms to train a high-quality automatic image recognition model. However, such an approach will bring data privacy and data availability challenges. To address these issues, we first present a semisupervised federated learning (SSFL) framework for privacy-preserving UAV image recognition. Specifically, we propose a model parameter mixing strategy to improve the naive combination of federated learning and semisupervised learning methods under two realistic scenarios (labels-at-client and labels-at-server), which is referred to as federated mixing (FedMix). Furthermore, there are significant differences in the number, features, and distribution of local data collected by UAVs using different camera modules in different environments, i.e., statistical heterogeneity. To alleviate the statistical heterogeneity problem, we propose an aggregation rule based on the frequency of the client’s participation in training, namely, the FedFreq aggregation rule, which can adjust the weight of the corresponding local model according to its frequency. Numerical results demonstrate that the performance of our proposed method is significantly better than those of the current baseline and is robust to different non-independent and identically distributed(IID) levels of client data.

Automatic Recognition of Leukemia AML Using Evolutionary Vision

In this work, an evolutionary vision approach is used for the automatic recognition of AML leukemia images. Unlike common approaches using convolutional neural networks, in the presented model the feature extraction process is transparent. Moreover, the structure of the obtained solutions is amenable to interpretation by a human user, which is a significant advantage over automatic recognition approaches based on deep neural networks. Experimental results show that the evolutionary vision approach can obtain satisfactory results on the AML leukemia recognition problem.

Animal Detection and Counting from UAV Images Using Convolutional Neural Networks

In the last decade, small unmanned aerial vehicles (UAVs/drones) have become increasingly popular in the airborne observation of large areas for many purposes, such as the monitoring of agricultural areas, the tracking of wild animals in their natural habitats, and the counting of livestock. Coupled with deep learning, they allow for automatic image processing and recognition. The aim of this work was to detect and count the deer population in northwestern Serbia from such images using deep neural networks, a tedious process that otherwise requires a lot of time and effort. In this paper, we present and compare the performance of several state-of-the-art network architectures, trained on a manually annotated set of images, and use it to predict the presence of objects in the rest of the dataset. We implemented three versions of the You Only Look Once (YOLO) architecture and a Single Shot Multibox Detector (SSD) to detect deer in a dense forest environment and measured their performance based on mean average precision (mAP), precision, recall, and F1 score. Moreover, we also evaluated the models based on their real-time performance. The results showed that the selected models were able to detect deer with a mean average precision of up to 70.45% and a confidence score of up to a 99%. The highest precision was achieved by the fourth version of YOLO with 86%, as well as the highest recall value of 75%. Its compressed version achieved slightly lower results, with 83% mAP in its best case, but it demonstrated four times better real-time performance. The counting function was applied on the best-performing models, providing us with the exact distribution of deer over all images. Yolov4 obtained an error of 8.3% in counting, while Yolov4-tiny mistook 12 deer, which accounted for an error of 7.1%.

Analysis of fragments larger than 2 mm generated by a picosatellite fragmentation experiment

Satellite breakup models rely on laboratory tests and in-space collision observations; current models can match fragments distributions generated by traditional satellites but may need to be improved for small spacecraft and modern satellites employing new configurations and materials. In the last years, ground tests have been employed to assess the influence of dimensions, materials and internal configurations on fragments distributions and to define the limits of the current models. In this context, an impact test was performed at the impact facility of the University of Padova to characterize the fragmentation of a picosatellite mock-up; more than 7000 fragments were collected, classified and analysed with automatic image recognition algorithms. It was observed that the experimental characteristic length distribution is line with the prediction of the NASA SBM even for the smallest size classes, while the fragments shape distribution is strongly affected by the materials employed in the picosatellite manufacturing.The subset of the collected fragments larger than 2 mm was recently subjected to a more detailed analysis: each fragment was individually weighed, and its three main dimensions were measured. In this paper, resulting fragments distributions are compared with literature data and the NASA Standard Breakup Model; in particular, an analytic relation between fragments characteristic length and size is found. In addition, results show that characteristic length and area-to-mass distributions are affected by the target materials and are clearly influenced by the size resolution of the analysed fragments.

Degree of Accuracy With Which Deep Learning for Ultrasound Images Identifies Osteochondritis Dissecans of the Humeral Capitellum

Background: Medical screening using ultrasonography (US) has been performed on young baseball players for early detection of osteochondritis dissecans (OCD) of the humeral capitellum. Deep learning (DL) and artificial intelligence (AI) techniques are widely adopted in the medical imaging research field. Purpose/Hypothesis: The purpose of this study was to calculate the diagnostic accuracy using DL for US images of OCD. We hypothesized that using DL for US imaging would improve the prediction accuracy of OCD. Study Design: Cohort study (Diagnosis); Level of evidence, 2. Methods: A total of 40 elbows (mean age of patients, 12.1 years) that were suspected of having OCD at a medical checkup and later confirmed by radiographs and magnetic resonance imaging were included in the study. The affected elbows were used as the OCD group and the contralateral elbows as the control group. From US videos, 100 images per elbow were captured from different angles, and 4000 images of the elbows were prepared for both groups. Of these, 80% were randomly selected by DL models and used as training data; the remaining were used as test data. Transfer learning was conducted using 3 pretrained DL models. The confusion matrix and the area under the receiver operating characteristic curve (AUC) were used to evaluate the model, and the visualization of the areas deemed important by the DL models was also performed. Furthermore, OCD regions were detected using an automatic image recognition model based on DL. Results: Classification of the OCD image by the DL model was performed; the best accuracy score was 0.87; the recall was 1.00. AUC was high for all DL models. Visualization of important features showed that AI predicted the presence of OCD by focusing on the irregularity or discontinuity of the surface of subchondral bone. In the detection of OCD task, the mean average precision was 0.83. Conclusion: The DL on US images identified OCD with high accuracy. The important features detected by the DL models correspond to the areas used by clinicians in screening the US images. The OCD was also detected with high accuracy using the object detection model. The AI model may be used in medical screening for OCD.

Retracted: Automatic Target Recognition of SAR Images Using Collaborative Representation

Retracted: Automatic Target Recognition of SAR Images Using Collaborative Representation

Method for the automatic recognition of cropland headland images based on deep learning

Method for the automatic recognition of cropland headland images based on deep learning

Fine-Grained Pornographic Image Recognition with Multi-Instance Learning

Image has become an essential medium for expressing meaning and disseminating information. Many images are uploaded to the Internet, among which some are pornographic, causing adverse effects on public psychological health. To create a clean and positive Internet environment, network enforcement agencies need an automatic and efficient pornographic image recognition tool. Previous studies on pornographic images mainly rely on convolutional neural networks (CNN). Because of CNN’s many parameters, they must rely on a large labeled training dataset, which takes work to build. To reduce the effect of the database on the recognition performance of pornographic images, many researchers view pornographic image recognition as a binary classification task. In actual application, when faced with pornographic images of various features, the performance and recognition accuracy of the network model often decrease. In addition, the pornographic content in images usually lies in several small-sized local regions, which are not a large proportion of the image. CNN, this kind of strong supervised learning method, usually cannot automatically focus on the pornographic area of the image, thus affecting the recognition accuracy of pornographic images. This paper established an image dataset with seven classes by crawling pornographic websites and Baidu Image Library. A weakly supervised pornographic image recognition method based on multiple instance learning (MIL) is proposed. The Squeeze and Extraction (SE) module is introduced in the feature extraction to strengthen the critical information and weaken the influence of non-key and useless information on the result of pornographic image recognition. To meet the requirements of the pooling layer operation in Multiple Instance Learning, we introduced the idea of an attention mechanism to weight and average instances. The experimental results show that the proposed method has better accuracy and F1 scores than other methods.

Measure of Similarity between GMMs Based on Geometry-Aware Dimensionality Reduction

Gaussian Mixture Models (GMMs) are used in many traditional expert systems and modern artificial intelligence tasks such as automatic speech recognition, image recognition and retrieval, pattern recognition, speaker recognition and verification, financial forecasting applications and others, as simple statistical representations of underlying data. Those representations typically require many high-dimensional GMM components that consume large computing resources and increase computation time. On the other hand, real-time applications require computationally efficient algorithms and for that reason, various GMM similarity measures and dimensionality reduction techniques have been examined to reduce the computational complexity. In this paper, a novel GMM similarity measure is proposed. The measure is based on a recently presented nonlinear geometry-aware dimensionality reduction algorithm for the manifold of Symmetric Positive Definite (SPD) matrices. The algorithm is applied over SPD representations of the original data. The local neighborhood information from the original high-dimensional parameter space is preserved by preserving distance to the local mean. Instead of dealing with high-dimensional parameter space, the method operates on much lower-dimensional space of transformed parameters. Resolving the distance between such representations is reduced to calculating the distance among lower-dimensional matrices. The method was tested within a texture recognition task where superior state-of-the-art performance in terms of the trade-off between recognition accuracy and computational complexity has been achieved in comparison with all baseline GMM similarity measures.

Evolution of morphology and microstructure of coarsened nanoporous gold studied by automatic thresholding and image recognition algorithms

The morphology and microstructure parameters of nanoporous gold (NPG) during thermal coarsening were investigated by automatic thresholding and image recognition algorithms. The ligaments width distributions, pore width distributions and curvature distributions are determined to quantitatively characterize microstructure of NPG. The results show a clear temperature dependence of these parameters. The measured activation energy (∼100 kJ/mol) for coarsening process supports the surface diffusion mechanism during ligaments coarsening of NPG. It was also found that the ligament and pore size are topological equivalent. Because of the dangling ligaments formed by pinch-off process have convex curvature, the ratio of positive curvature is higher than that of negative curvature. This structural geometry and microstructure characteristics analysis of NPG could help researchers better understanding the structure evolution, physical and chemical properties.