Image Recognition Research Articles

Comparative study of teachable machine for forest fire and smoke detection by drone

Forests play a vital role in maintaining ecological equilibrium and serving as vital habitats for wildlife. They regulate global climate, safeguard soil and water resources, and provide crucial ecosystem services such as air and water purification, essential for human well-being and sustainable development. Forest fires wreak havoc on ecosystems and wildlife, emitting harmful pollutants, disrupting communities, and increasing the risk of erosion and landslides. Detecting forest fires through satellite imaging, aerial reconnaissance, and ground-based sensors is pivotal for early detection and containment, safeguarding human lives, wildlife, and preserving natural resources for future generations. Utilizing drones and deep learning (DL) algorithms can significantly enhance early fire detection and minimize their devastating impact. In this paper, we examine teachable machine, a Google tool for creating DL models. We compare the top model generated by teachable machine for fire and smoke detection to models obtained through transfer learning from established DL models in image recognition and computer vision (CV), such as VGG16, VGG19, MobileNet, MobileNetv2, and MobileNetv3. The results underscore the significance of employing the teachable machine model in specific fire and smoke detection scenarios.

TQU-HG dataset and comparative study for hand gesture recognition of RGB-based images using deep learning

Hand gesture recognition has great applications in human-computer interaction (HCI), human-robot interaction (HRI), and supporting the deaf and mute. To build a hand gesture recognition model using deep learning (DL) with high results then needs to be trained on many data and in many different conditions and contexts. In this paper, we publish the TQU-HG dataset of large RGB images with low resolution (640×480) pixels, low light conditions, and fast speed (16 fps). TQU-HG dataset includes 60,000 images collected from 20 people (10 male, 10 female) with 15 gestures of both left and right hands. A comparative study with two branches: i) based on Mediapipe TML and ii) Based on convolutional neural networks (CNNs) (you only look once (YOLO); YOLOv5, YOLOv6, YOLOv7, YOLOv8, YOLO-Nas, single shot multiBox detector (SSD) VGG16, residual network (ResNet)18, ResNext50, ResNet152, ResNext50, MobileNet V3 small, and MobileNet V3 large), the architecture and operation of CNNs models are also introduced in detail. We especially fine-tune the model and evaluate it on TQU-HG and HaGRID datasets. The quantitative results of the training and testing are presented (F1-score of YOLOv8, YOLO-Nas, MobileNet V3 small, ResNet50 is 98.99%, 98.98%, 99.27%, 99.36%, respectively on the TQU-HG dataset and is 99.21%, 99.37%, 99.36%, 86.4%, 98.3%, respectively on the HaGRID dataset). The computation time of YOLOv8 is 6.19 fps on the CPU and 18.28 fps on the GPU.

Enhancing accessibility and discoverability of digital archive images through automated image recognition tool

This research paper presents a comprehensive evaluation of the effectiveness of Imagga and Google cloud vision application programming interface (API) as image recognition tools for generating metadata in digital archive images. The assessment encompasses a diverse range of archive images, including those without text, images with text, and both color and black-and-white images. Through the use of evaluation metrics such as cosine similarity, word overlap similarity, recall, precision, and F1 score, the performance of these tools is quantitatively measured. The findings highlight the strong individual performance of both Imagga and Google cloud vision API, with the combined metadata outputs achieving significantly higher scores across all metrics. This emphasizes the potential benefits of employing a combined approach, leveraging the strengths of multiple tools to enhance the reliability and robustness of the metadata extraction process. The findings contribute to the advancement of metadata management in digital archives and underscore the importance of utilizing multiple tools for improved performance in image metadata generation.

Malaria cell identification using improved machine learning and modified deep learning architecture

Malaria continues to be a serious problem for public health because of its occurrence in tropical and subtropical areas with inadequate healthcare systems and few resources. For prompt intervention and treatment of malaria, effective and precise diagnosis is essential. Professional pathologists examine blood smear films by hand to get a microscopic diagnosis and another way they will do a rapid antigen malaria test which produces the result of 50% accuracy. Convolutional neural network (CNN) is a type of deep learning (DL) model that has been effectively used for a variety of image recognition applications. Our suggested approach uses, improved machine learning (IML) methods like support vector machine (SVM)+principal component analysis (PCA) fit, SVM+t-distributed stochastic neighbor embedding (t-SNE) fit, and CNN architecture with an accuracy of 86.23%, 88.27%, and 97.16% accuracy respectively, to combine feature extraction, data augmentation, and modify the layers by including the SVM algorithm in the final layer of the CNN architecture. The proposed method will significantly reduce pathologists' burden by automating the identification of malaria and improving diagnosis accuracy in resourceconstrained contexts

Resource-efficient federated learning over IoAT for rice leaf disease classification

Rice is an important staple food in Asia. It is produced and consumed in large quantities. It contributes to 15% of protein intake and 21% of total per capita energy intake in the region, underscoring its important role as a primary global food source. Conversely, rice plants are heavily affected by bacterial, fungal and other microbial diseases, resulting in reduced plant health and crop yields and posing a major challenge for rice farmers. Manual diagnosis of these diseases is particularly problematic in regions with a shortage of agricultural experts. Farmers with insufficient experience sometimes incorrectly identify these diseases by hand. Recent advances in deep learning (DL) models offer a promising solution through automatic image recognition systems that can be very helpful in accurately identifying these diseases. This manuscript presents a resource-efficient federated learning IoAT (Internet of Agriculture Things) approach for rice leaf disease classification. This approach incorporates two key strategies, namely federated transfer learning and feature extraction, and evaluates their performance on various metrics such as accuracy, loss, precision, recall, AUC, and resource-related parameters such as GPU memory consumption, virtual memory, CPU and GPU process utilization. The dataset used in the study includes 5932 images of rice leaf diseases categorized as bacterial leaf blight, blast, brown spot and tungro. The research investigates the application of federated transfer learning and federated feature extraction techniques to classify rice leaf disease images. It performs a comparative analysis of their performance and resource utilization. EfficientNetB3, with an impressive validation accuracy of 99%, is identified as the base model for the federated learning (FL) environment. Furthermore, we implement FL using both transfer learning and feature extraction methods and compare their results on a number of performance and resource related parameters. It is shown that the federated feature extraction approach has lower GPU memory and process utilization compared to the federated transfer learning approach and is thus more resource efficient. Extracting features before model training results in a lightweight model, which is especially beneficial for FL, IoAT devices that require efficient execution on edge devices. Therefore, the proposed approach is presented as a lightweight, privacy-friendly and resource-efficient solution for rice leaf disease classification.

Methodology for eliminating plain regions from captured images

Finding relevant content and extracting information from images is highly significant. Still, it may be challenging to do so because of changes within the textual contents, such as typefaces, size, line orientation, sophisticated backgrounds in images, and non-uniform illuminations. Despite these challenges, extracting content from captured images is still very important. Proficient textual content image recognition abilities extract text from the images to get over these issues. Despite the availability of several optical character recognition (OCR) techniques, this issue has yet to be resolved. Captured images with text are a rich source of information that should be presented so that viewers may make informed decisions. Because of this, it has become a complicated process to extract the text from an image because the text might be of poor quality, has a variety of fonts and styles, and occasionally have a complicated backdrop, among other things. Several approaches have been tried. However, finding a solution remains challenging. The maximally stable external regions (MSER) approach is developed to identify the text region in a picture. MSER is utilized to elevate the plain regions outside the text and non-text areas using geometric features and stroke width variation qualities.

Enhanced Virtual Try on for Clothing using Deep learning

Abstract: An increasing number of individuals are shopping online, particularly for clothing. However, while purchasing online, actual try-ons are not allowed, which limits the customer’s ability to see how a garment would appear on them. To enable buyers to virtually try on the desired clothing from an online store, an imagebased try-on system was devised. The ability to practically put on several clothes has increased customer interest in shopping. Image Recognition and Feature Extraction was the important part of the model to get desired outcomes. And To address these problems, we implement the CP-VTON+ (Clothing shape and texture Preserving VTON), which performs substan- tially better than the most advanced in sense of both quantitative and qualitative evaluations. The CP-VTON+ model not only adapt to detangle clothes from the person, but also maintains the clothing’s details on the try-on outcomes.

AI-Powered Autonomous Checkout System for Retail Stores

Abstract: Auto Bill offers a smart, automated checkout service for shops, leveraging state-of-the-art machine learning and computer vision technologies to provide an exceptional shopping experience. Designed to minimize interactions for security purposes during the pandemic, the innovation lies in utilizing modern techniques like image processing and weight sensors to create a novel shopping experience. This system aims to surpass traditional weight scales by incorporating advanced image recognition technologies, benefiting customers with a highly user-friendly interface, reduced checkout times, and improved accessibility. With a multifaceted approach involving deep learning neural networks and a web-based interface for transactions and inventory monitoring, Auto Bill envisions a future where retail checkout is taken to new heights. Supermarkets adopting this system can enhance customer service throughits advanced machine learning and camera technology, offering a seamless shopping experience. By integrating easily with existing POS systems, this solution streamlines operations and improves overall efficiency in the retail industry. With continuous testing and adjustments, Auto Bill aims to refine retail checkout to unprecedented levels of accuracy and customer satisfaction.

Pet Classification Using Transfer Learning

Abstract: This study delves into applying transfer learning to pet classification using pre-trained neural networks, specifically convolutional neural networks (CNNs). It adapts models originally trained on ImageNet to distinguish between different pet species. By fine-tuning these models on a custom pet dataset, the study aims to leverage previous knowledge to improve classification accuracy, even with limited labeled data. Various transfer learning strategies, model architectures, and hyper parameters are analyzed to identify the most effective configuration for accurate and efficient pet classification. The research has implications for pet monitoring, identification, and healthcare, contributing to computer vision and deep learning in specialized image recognition.

Botanic Shield: A Deep Learning Solution for Early Detection of Plant Disease

Abstract: Plant diseases threaten both food security and the botanical system of the environment which canlead to low- grade food quality and life style. Early detection of plants diseases is the key feature in plantproduction which increase the better growth of plant, results in best yield and better food quality andquantity. The plant condition can be determined byusing the photos of the plant which will be in the formof jpg using digital image processing and the current emerging technology Deep Learning. DeepLearning has made breakthroughs in the field of digital image processing, far superior to traditional methods. The Digital image processing process happens through convolution neural network (CNN) and computer vision techniques, the image recognition technology based on deep learning does not need to extractspecificfeatures, and only through iterative learning can find appropriate features, which can acquire global andcontextual features of images, and has strong robustness and higher recognition accuracy. By inputting a test image into the classification network, the network analyses the input image and returns a label that classifiesthe image. The added feature is thatit also suggeststhe required pesticides after the detection

How to Reduce the Number of Inspectors: Operation Mode and Modeling Simulation of Centralized Image Recognition in Railway Passenger Security Inspection

Security inspection is an essential step for passengers entering railway stations in China. Under the existing security inspection operation mode, each X-ray machine needs an image inspector to recognize the X-ray image, which poses the issue of resource optimization for inspectors. Based on the networking of security inspection X-ray machines, an innovative centralized image recognition (CIR) operation mode for railway security inspections is proposed, and it is worth noting that network and device failures are risk factors for this mode. The method for assigning image recognition personnel is designed to reduce labor force needed for security inspection. Using a combination of agent-based modeling method and computer simulation technology, the four elements of passengers, security inspection X-ray machines, X-ray image inspectors, and secondary baggage inspection personnel are modeled, and CIR indicators to study the effectiveness of the operation are designed. Based on the real departing passenger data of the Zhangjiakou section of the Beijing Zhangjiakou Line in China, simulation is conducted in two scenarios: large passenger flow and small passenger flow. The research is conducted from three aspects: operational efficiency, travel experience, and system complexity. Experiments have shown that, with a slight increase in passenger waiting time, the average number of X-ray image inspectors is reduced by more than 70% compared with the existing operation mode. The best pick-up location for passengers is related to passenger flow. The proposed model can analyze the effectiveness of CIR operation mode and provide decisive support for the application of CIR operation mode.

Bismuth-based ferroelectric memristive device induced by interface barrier for neuromorphic computing

The development of the Internet of Things (IoT) not only facilitates our lives but also dramatically grows data. Artificial synaptic devices appear to represent an emerging physical solution compared to the existing computational architectures. Memristive devices are of great interest with their high integration and low power consumption in the field of synaptic devices. In this work, we demonstrate short-term plasticity (STP) and long-term plasticity (LTP) by regulating synaptic weights with different stimulus pulses. Moreover, the application in neuromorphic computing is further exhibited by image recognition with an accuracy of 95.2 % under the modified National Institute of Standards and Technology database. Therefore, Nd-doped Bi4Ti3O12 memristors, as emerging artificial synaptic devices, are expected to achieve breakthroughs in artificial intelligence and promote the development of neuromorphic computing and intelligent systems.

Deep Learning-Based Surgical Treatment Recommendation and Nonsurgical Prognosis Status Classification for Scaphoid Fractures by Automated X-ray Image Recognition

Biomedical information retrieval for diagnosis, treatment and prognosis has been studied for a long time. In particular, image recognition using deep learning has been shown to be very effective for cancers and diseases. In these fields, scaphoid fracture recognition is a hot topic because the appearance of scaphoid fractures is not easy to detect. Although there have been a number of recent studies on this topic, no studies focused their attention on surgical treatment recommendations and nonsurgical prognosis status classification. Indeed, a successful treatment recommendation will assist the doctor in selecting an effective treatment, and the prognosis status classification will help a radiologist recognize the image more efficiently. For these purposes, in this paper, we propose potential solutions through a comprehensive empirical study assessing the effectiveness of recent deep learning techniques on surgical treatment recommendation and nonsurgical prognosis status classification. In the proposed system, the scaphoid is firstly segmented from an unknown X-ray image. Next, for surgical treatment recommendation, the fractures are further filtered and recognized. According to the recognition result, the surgical treatment recommendation is generated. Finally, even without sufficient fracture information, the doctor can still make an effective decision to opt for surgery or not. Moreover, for nonsurgical patients, the current prognosis status of avascular necrosis, non-union and union can be classified. The related experimental results made using a real dataset reveal that the surgical treatment recommendation reached 80% and 86% in accuracy and AUC (Area Under the Curve), respectively, while the nonsurgical prognosis status classification reached 91% and 96%, respectively. Further, the methods using transfer learning and data augmentation can bring out obvious improvements, which, on average, reached 21.9%, 28.9% and 5.6%, 7.8% for surgical treatment recommendations and nonsurgical prognosis image classification, respectively. Based on the experimental results, the recommended methods in this paper are DenseNet169 and ResNet50 for surgical treatment recommendation and nonsurgical prognosis status classification, respectively. We believe that this paper can provide an important reference for future research on surgical treatment recommendation and nonsurgical prognosis classification for scaphoid fractures.

Thermal radiation image recognition camera using target detection techniques with human computer interaction

Backgroundand Purpose: The evolution of object detection methodologies has shifted from conventional approaches to modern deep learning architectures to overcome limitations in handling object variations. The goal was to introduce YOLO, a groundbreaking architecture enabling real-time object detection, leading to the development of an AI camera using YOLO for thermal imaging object recognition. This innovation aimed to precisely identify animals, humans, and plants in thermal images while providing a user-friendly experience. MethodologyTraditional methods' limitations were highlighted, emphasizing YOLO's single-pass architecture and its integration into the AI camera system, aligning with Human-Computer Interaction principles. This AI camera system can detect object between Animals, Humans and plants. Stratified K-Fold Cross-Validation was utilized to refine the AI model, ensuring accuracy and generalization by representing object classes proportionally. ResultsEmploying Stratified K-Fold Cross-Validation achieved an overall 97.55% detection accuracy and 85.75% sensitivity, showcasing the AI model's robustness. Comparative analysis with R-CNNs revealed YOLO's 18% accuracy improvement, affirming its superiority and strategic implementation in the AI camera system. These findings validated YOLO's efficiency in thermal image object detection. ConclusionThe conclusion highlights the profound impact of YOLO, specifically within the domain of thermal image recognition, substantiating its incorporation into the AI camera system in adherence to Human-Computer Interaction (HCI) principles. The commendable precision of YOLO, coupled with its real-time capabilities and user-centric design, has resulted in the development of an advanced thermal image object detection system, representing a notable advancement in the field of object recognition technology.

Endangered Bird Species Classification using Machine Learning

The conservation of endangered bird species is a critical aspect of biodiversity preservation. Traditional methods of identifying and classifying these species are often labor-intensive and time-consuming. In recent years, advancements in machine learning have offered promising alternatives for enhancing the accuracy and efficiency of such tasks. This paper explores the application of various machine learning algorithms to the classification of endangered bird species. By leveraging a dataset comprising images and audio recordings of bird calls, we train and evaluate models such as Convolutional Neural Networks (CNNs) for image recognition and Recurrent Neural Networks (RNNs) for audio classification. Our results demonstrate that these models can achieve high accuracy rates, significantly surpassing traditional methods. Furthermore, we discuss the importance of feature selection, data augmentation, and the integration of multimodal data in improving model performance. The findings underscore the potential of machine learning to revolutionize wildlife conservation efforts, providing a scalable and robust tool for the timely identification and protection of endangered bird species.

Verinote - Fake Currency Detection Using Convolutional Neural Network

The proliferation of counterfeit money poses a significant threat to financial systems and economies worldwide. To solve this problem, advanced technological solutions have emerged, such as the counterfeit money detection system “FCDS”. The system leverages advanced image processing, machine learning and data analysis techniques to identify counterfeit bills accurately and effectively [1]. FCDS works by analyzing various security features found on legal tender, including watermarks, security chains, holograms and microprinting. Using image recognition and pattern analysis, the system distinguishes between real and fake money. Machine learning algorithms play a central role in training systems to recognize the subtle nuances that counterfeiters try to reproduce. FCDS can be deployed in a variety of contexts, from banks and financial institutions to retail businesses, providing a robust and scalable solution to combat counterfeiting [1]. By quickly identifying fraudulent notes, it helps prevent economic loss and maintain the integrity of financial transactions. This summary describes the nature of counterfeit currency detection systems and its importance in maintaining financial security and confidence. Its integration into modern banking and commerce systems represents an important step towards a counterfeit-proof financial ecosystem.

A new novel recognition and positioning system of black light-absorbing volute for automation depalletizing development

Abstract In the application of vision measurement, the black light-absorbing object is difficult to reflect the structured light from infrared emitter of the RGB-D camera. Therefore, an image recognition algorithm based on reference environment information is proposed to acquire the spatial positioning information of black volutes in the depalletizing system. The hardware system of the depalletizing system is mainly constructed of an upper computer, a six-axis industrial robot, an RGB-D camera and an end adsorption device. Firstly, the horizontal position information of each volute placed on the cardboard is obtained by the depth differences between the cardboard and the volute. Then, the depth information of the volute is obtained by the upper cardboard depth through collecting the position of the end vacuum suction cup triggered by feedback signal from vacuum generator. Secondly, a regional planar hand-eye calibration method is developed to improve the calibration accuracy in two-dimensional coordinates. The regional calibration method divides the robot working area into four regions: upper left, lower left, upper right, and lower right. The transformation matrix of each region is calculated separately. Finally, the depalletizing experiment is conducted on the three types of volutes. It is concluded that the average positioning error of the grasping center point of each volute obtained by our method is 3.795 mm, and its standard deviation is 1.769 mm. The average value of regional planar hand-eye calibration error is 4.044 mm, and its standard deviation is 1.501 mm. Under a stack of materials with dimensions of 1350 mm × 1350 mm × 1500 mm, the maximum error is controlled within 15mm. Additionally, when combined with the end feedback compensation mechanism, the success rate for grasping all three volutes reaches 100%.

Research on Improved Road Visual Navigation Recognition Method Based on DeepLabV3+ in Pitaya Orchard

Traditional DeepLabV3+ image semantic segmentation methods face challenges in pitaya orchard environments characterized by multiple interference factors, complex image backgrounds, high computational complexity, and extensive memory consumption. This paper introduces an improved visual navigation path recognition method for pitaya orchards. Initially, DeepLabV3+ utilizes a lightweight MobileNetV2 as its primary feature extraction backbone, which is augmented with a Pyramid Split Attention (PSA) module placed after the Atrous Spatial Pyramid Pooling (ASPP) module. This improvement enhances the spatial feature representation of feature maps, thereby sharpening the segmentation boundaries. Additionally, an Efficient Channel Attention Network (ECANet) mechanism is integrated with the lower-level features of MobileNetV2 to reduce computational complexity and refine the clarity of target boundaries. The paper also designs a navigation path extraction algorithm, which fits the road mask regions segmented by the model to achieve precise navigation path recognition. Experimental findings show that the enhanced DeepLabV3+ model achieved a Mean Intersection over Union (MIoU) and average pixel accuracy of 95.79% and 97.81%, respectively. These figures represent increases of 0.59 and 0.41 percentage points when contrasted with the original model. Furthermore, the model’s memory consumption is reduced by 85.64%, 84.70%, and 85.06% when contrasted with the Pyramid Scene Parsing Network (PSPNet), U-Net, and Fully Convolutional Network (FCN) models, respectively. This reduction makes the proposed model more efficient while maintaining high segmentation accuracy, thus supporting enhanced operational efficiency in practical applications. The test results for navigation path recognition accuracy reveal that the angle error between the navigation centerline extracted using the least squares method and the manually fitted centerline is less than 5°. Additionally, the average deviation between the road centerlines extracted under three different lighting conditions and the actual road centerline is only 2.66 pixels, with an average image recognition time of 0.10 s. This performance suggests that the study can provide an effective reference for visual navigation in smart agriculture.

Dynamic Failure Characteristics of Sandstone Containing Different Angles of Pre-Existing Crack Defects

Fracture within the rock is one of the main factors leading to rock destabilization and has a significant effect on the stability of the project. In this study, sandstone is used as a research target, specimens with crack inclination angles of 0°, 30°, 45°, 60°, and 90° are prefabricated, and the split Hopkinson pressure bar (SHPB) impact test of sandstone with cracks is carried out based on digital image recognition technology to explore the dynamic damage characteristics of the specimens with five angles. The basic mechanical parameters of sandstone are tested to determine the RHT model intrinsic parameters, and the numerical computational RHT model of sandstone containing crack defects is established, which is verified in comparison with the test to analyze the validity of the model. Finally, the failure characteristics of the numerical model under initial stress were carried out. The study shows the following: with the increase in the fracture angle, the dynamic compressive strength and deformation modulus are distributed in a slanting V-shape, and the inclination angle of 45° is the smallest. The strain rate and energy dissipation rate are distributed in a slanting N-shape, and the inclination angle of 45° is the largest. The transmittance shows a decreasing trend, which is the opposite of the reflectivity pattern. The crack angle determines the location and direction of the initial crack, which affects the failure mode. In addition, the parameters of the RHT constitutive model suitable for sandstone are obtained, and the damage and strength patterns of the established RHT model are highly consistent with the laboratory test results. The damage range of numerical models for crack defects with different inclination angles is negatively correlated with confining pressure values and positively correlated with axial pressure values. The damage zones are symmetrically distributed approximately perpendicular to the direction of cracks, and the confining pressure has a contributing role in the peak of the element stresses; however, the axial compression has no contribution in the peak of the element stresses.

Enhancement of Mine Images through Reflectance Estimation of V Channel Using Retinex Theory

The dim lighting and excessive dust in underground mines often result in uneven illumination, blurriness, and loss of detail in surveillance images, which hinders subsequent intelligent image recognition. To address the limitations of the existing image enhancement algorithms in terms of generalization and accuracy, this paper proposes an unsupervised method for enhancing mine images in the hue–saturation–value (HSV) color space. Inspired by the HSV color space, the method first converts RGB images to the HSV space and integrates Retinex theory into the brightness (V channel). Additionally, a random perturbation technique is designed for the brightness. Within the same scene, a U-Net-based reflectance estimation network is constructed by enforcing consistency between the original reflectance and the perturbed reflectance, incorporating ResNeSt blocks and a multi-scale channel pixel attention module to improve accuracy. Finally, an enhanced image is obtained by recombining the original hue (H channel), brightness, and saturation (S channel), and converting back to the RGB space. Importantly, this image enhancement algorithm does not require any normally illuminated images during training. Extensive experiments demonstrated that the proposed method outperformed most existing unsupervised low-light image enhancement methods, qualitatively and quantitatively, achieving a competitive performance comparable to many supervised methods. Specifically, our method achieved the highest PSNR value of 22.18, indicating significant improvements compared to the other methods, and surpassing the second-best WCDM method by 10.3%. In terms of SSIM, our method also performed exceptionally well, achieving a value of 0.807, surpassing all other methods, and improving upon the second-place WCDM method by 19.5%. These results demonstrate that our proposed method significantly enhanced image quality and similarity, far exceeding the performance of the other algorithms.