Application Of Image Processing Techniques Research Articles

Rice leaf disease detection using the Stretched Neighborhood Effect Color to Grayscale method and Machine Learning

The emergence of Machine Learning (ML) technologies and their integration into agriculture has demonstrated a significant impact on disease detection in crops, enabling continuous monitoring and enhancing risk planning and management. This study applied image processing techniques such as thresholding, gamma correction, and the Stretched Neighborhood Effect Color to Grayscale (SNECG) method, alongside ML, to develop a predictive model for identifying five types of rice diseases. The ML techniques used included Logistic Regression, Multilayer Perceptron, Support Vector Machines, Decision Trees, and Random Forests (RF). Hyperparameters were optimized and evaluated through 5-fold cross-validation. In the results, the SNECG method successfully converted images to grayscale, capturing essential features of lesions on rice leaves. The ML models developed with these techniques showed evaluation metrics exceeding 80%, with the RF model (precision = 88.31%) demonstrating superior performance. Additionally, the RF model was integrated into an interface designed for agricultural decision-making. The practical application of the developed model could significantly improve the ability to detect and manage diseases in rice crops.

Spatial and temporal deep learning algorithms for defect segmentation in infrared thermographic imaging of carbon fibre-reinforced polymers

ABSTRACT For non-destructive evaluation, the segmentation of infrared thermographic images of carbon fibre composites is a critical task in material characterisation and quality assessment. This paper presents a study on the application of image processing techniques, particularly adaptive thresholding, and advanced neural network models, including U-Net, DeepLabv3, and BiLSTM, for the segmentation of infrared images. This work introduces the innovative combination of DeepLabv3 and BiLSTM applied in infrared images of carbon fibre-reinforced polymer samples for the first time, proposing it as a novel approach for enhancing the accuracy of segmentation tasks. An experimental comparison of these models was conducted to assess their effectiveness in identifying artificial defects in these images. The performance of each model was evaluated using the F1-Score and Intersection over Union (IoU) metrics. The results demonstrate that the proposed combination of DeepLabv3 and BiLSTM outperforms other methods, achieving an F1-Score of 0.96 and an IoU of 0.83, showcasing its potential for advanced material analysis and quality control.

Novel and cost-effective CNC tool condition monitoring through image processing techniques

CNC machining is an important part of the manufacturing industry. This paper introduces a novel and efficient approach for tool condition monitoring in CNC machine operations through the application of image processing techniques. By utilizing a consumer-grade camera capable of recording videos at 60 frames per second, the study demonstrates a cost-effective method to detect tool breakage and identify edge fractures. Basic image processing techniques, including frame extraction, background subtraction, thresholding, and morphological operations, are applied to analyze captured images and videos. This research not only offers a practical solution to enhance the efficiency and accuracy of CNC machine operations but also aligns with advancements in smart manufacturing and Industry 4.0. The findings will showcase the proposed system’s effectiveness in actual CNC machine environments, underscoring its potential to enhance maintenance strategies and operational productivity for small and medium-sized manufacturers without incurring high costs. Furthermore, it sets the stage for future investigations in this domain, indicating the possibility for enhancements through machine learning and an expanded application of these monitoring techniques.

An Article Review on Vision-Based Defect Detection Technologies for Reinforced Concrete Bridges

Bridges are crucial and the most vulnerable element in the infrastructure systems. A major challenge is to maintain bridge structures at a sufficient level of safety. Scheduled inspections in these structures are important to prevent any failure. The requirement of periodic inspection is urgently needed to maintain the bridges in safe operating condition for the public. Visual inspection is currently the main form for the flaw’s inspection. Nevertheless, it is suffering from time consuming and some limitations related to subjectivity and uncertainty. Due to the complexity of bridge structure, automatic defect detection is an urgent requirement for reinforced concrete bridges. In view of this, the creation and utilization of computer vision method has received considerable attention in several applications of civil engineering. Thus, this paper introduces a comprehensive study in computer vision-based defect detection related to concrete bridges. In this study, a detailed survey is undertaken to identify the research problems and the accomplishments to date in this field. Accordingly, 50 studies between peer-reviewed publications and conference papers Scopus found in are reviewed. Through the analysis, the current review divided the image technology into three groups based on: 1) image processing; 2) machine learning; and 3) quantifying the severity of defects by identifying their parameters. This article highlights the difference and the advantages and disadvantages of applying image processing techniques and machine learning. The paper identifies the types of defects detected by image technology in previous studies and their shortcomings in determining some parameters related to those defects. Finally, this research addresses issues related to the efficiency of detection and the main factors to be considered that may help further research in image-based approaches for defect detection effectively in concrete bridges.

A Comprehensive Review on Accident Detection Using Artificial Intelligence and Machine Learning

This paper reviews the role of Artificial Intelligence (AI) and Machine Learning (ML) in accident detection systems, focusing on the application of image processing techniques and Convolutional Neural Networks (CNNs) to enable detection. Through a comprehensive analysis of recent methodologies, challenges, limitations, and future research directions, this paper aims to provide a detailed understanding of how AI and ML are advancing the field of accident detection

Blind Recognition of Frame Synchronization Based on Deep Learning.

In this paper, a deep-learning-based frame synchronization blind recognition algorithm is proposed to improve the detection performance in non-cooperative communication systems. Current methods face challenges in accurately detecting frames under high bit error rates (BER). Our approach begins with flat-top interpolation of binary data and converting it into a series of grayscale images, enabling the application of image processing techniques. By incorporating a scaling factor, we generate RGB images. Based on the matching radius, frame length, and frame synchronization code, RGB images with distinct stripe features are classified as positive samples for each category, while the remaining images are classified as negative samples. Finally, the neural network is trained on these sets to classify test data effectively. Simulation results demonstrate that the proposed algorithm achieves a 100% probability in frame recognition when BER is below 0.2. Even with a BER of 0.25, the recognition probability remains above 90%, which exhibits a performance improvement of over 60% compared with traditional algorithms. This work addresses the shortcomings of existing methods under high error conditions, and the idea of converting sequences into RGB images also provides a reliable solution for frame synchronization in challenging communication environments.

Performance Analysis of a Color-Code-Based Optical Camera Communication System

In this study, we present a visible light communication (VLC) system that analyzes the performance of an optical camera communication (OCC) system, utilizing a mobile phone camera as the receiver and a computer monitor as the transmitter. By creating color channels in the form of a 4 × 4 matrix within a frame, we determine the parameters that affect the successful transmission of data packets. Factors such as the brightness or darkness of the test room, the light color of the lamp in the illuminated environment, the effects of daylight when the monitor is positioned in front of a window, and issues related to dead pixels and light bleed originating from the monitor’s production process have been considered to ensure accurate data transmission. In this context, we utilized the PyCharm, Pydroid, Python, Tkinter, and OpenCV platforms for programming the transmitter and receiver units. Through the application of image processing techniques, we mitigated the effects of daylight on communication performance, thereby proposing a superior system compared to standard VLC systems that incorporate photodiodes. Additionally, considering objectives such as the maximum number of channels and the maximum distance, we regulated the sizes of the channels, the distances between the channels, and the number of channels. The NumPy library, compatible with Python–Tkinter, was employed to determine the color levels and dimensions of the channels. We investigate the effects of RGB and HSV color spaces on the data transmission rate and communication distance. Furthermore, the impact of the distance between color channels on color detection performance is discussed in detail.

Graphene nanoplatelet additive impact on compression and bending strength of S-2 glass face-sheet honeycomb sandwich panels: An experimental study

In this study, the three-point bending, edge-wise, and flat-wise strengths of Nomex® sandwich structures with un-doped S-2 glass fiber face-sheet (Neat/S-2/glass) and those with Graphene nanoplatelet-doped S-2 glass fiber face-sheet (GNPs/S-2/glass) were examined parametrically. In the macroscopic images obtained post-experimentation, the damage to the structures’ face layer, core, and interface with GNP (graphene nanoplatelet) additives was visualized. Morphological images depicting the damaged structures were obtained by applying image processing techniques to the macroscopic images. Finally, the fracture mechanisms of fiber/fiber and fiber/matrix interfaces within the shell layers were examined in SEM images. As a result of the study, the addition of GNPs increased the strength of the interfaces between the fiber layers and the fiber core within the face layers of S-2 glass face-sheet Nomex® sandwich structures, leading to a 40% enhancement in bending strength, a 14% improvement in core compressive strength, and a 2.85% increase in edge-wise compressive strength. Moreover, the delamination damage between the face layer and the core, which undermines the structural integrity of the sandwich structure, was not observed in GNPs/S-2/glass sandwich structures. Thus, the damage propagation throughout the sandwich structure contributed to enhancing its load-carrying capacity. The collapse extent of the sandwich structures, determined through image processing, measured 8676 mm in Neat/S-2/glass sandwiches, whereas it amounted to 10,410 mm in GNPs/S-2/glass sandwiches. The inclusion of GNPs additive augmented the collapse strength of the sandwich structure by 19.9%. Utilizing GNPs as a filler in the matrix material has been recognized as a practical approach to enhancing the mechanical strength of sandwich structures.

Key parameters and effects in image processing and aggregate–aggregate contact calculation of asphalt mixtures

At present, there is no standard way of acquiring images, addressing errors in various image processing steps, and analyzing skeleton contact parameters. To promote the application of image processing technique in the design, construction monitoring, and quality inspection of asphalt mixtures, this study investigates the key factors affecting the two-dimensional image processing accuracy of asphalt mixtures, and the impact of parameters on the contact structure of asphalt mixture skeleton during the contact analysis is studied. Finally, data accuracy requirements for image processing and contact analysis are proposed. The results indicate that the camera lens should be parallel to the slice with a tilt angle of less than 15°, the error of the passing rate of each sieve remains within 5.0 %. The smaller the aggregate particle size, the greater the impact of aggregate edge recognition error on the aggregate particle size, aggregates with a particle size greater than 2.36 mm are less affected by edge recognition error. For aggregate–aggregate contact analysis, it is appropriate to set the contact threshold to 0.23 times the minimum calculated particle size. To ensure the accuracy of image processing and contact analysis, at least 10 cross-sectional images should be prepared, and the average value can be taken as the representative value of the indices. The conclusions can improve the accuracy and reliability of image processing, providing more accurate guidance for the design and construction of asphalt mixture.

Application of Image Processing Techniques in Agricultural Remote Sensing UAVs

This study aims to explore the application of image processing techniques in agricultural remote sensing UAVs to improve the accuracy and usability of agricultural remote sensing data. Through literature review and empirical research, the current application status, advantages, and challenges of image processing techniques in agricultural remote sensing UAVs were systematically studied. The study first analyzed and summarized the principles and characteristics of image processing techniques, and then conducted case studies and field investigations on their specific applications in image preprocessing, feature extraction, and image classification. The research results show that image processing techniques have high accuracy and usability in agricultural remote sensing UAVs, enabling the extraction of agricultural features and identification of crop diseases. However, challenges such as complex data processing, diverse method selection, and algorithm optimization still exist in the application of image processing techniques in agricultural remote sensing UAVs. The findings of this study have important implications for guiding the application and promotion of image processing techniques in agricultural remote sensing UAVs.

Identification of dragon trees and fruits in ham Thuan Bac growing areas, Phan Thiet city, Binh Thuan province, Vietnam

With the development of Computer Vision, we can effectively and accurately identify trees, fruit or object images. But to build a high-performance image dataset for tree identification problems in Agriculture is a challenge. Realizing that Vietnam is a country with strong agriculture with many tropical fruits grown widely such as Dragon fruit, Mangosteen, Mango, Orange, Lychee, Longan … We chose the Dragon Fruit tree for the data set. of my proposed images, all images will be collected using the close-up method, including tasks such as taking photos of Dragon Fruit trees from many angles and in different conditions (weather, temperature, light, …). In this article, we want to improve the data quality of the collected images so we have applied image processing techniques such as noise filtering (using Gaussian filter), image quality enhancement (image rotation), flip the image, zoom out, zoom in, etc.). From the collected Dragon Fruit tree data set, we will propose to use the Faster R–CNN model for this data set to build a tree and Dragon Fruit identification system.

E-Citrus: A Cloud-Based Citrus Pest and Disease Detection, Diagnostic and Prevention using Convolutional Neural Network

Citrus fruit yields in the Philippines have been fluctuating dramatically in recent years. Diseases, pests, and soil inadequacy have all contributed to the citrus industry's severe decline. More than 15 viruses and virus-like diseases have infected Citrus. Agricultural productivity must improve for a country to be progressive. Resources should be utilized to their full potential, diseases and pests should be controlled efficiently, and technological advancements must be adopted. This application will identify and map common pests and diseases of citrus fruits in Oriental Mindoro, apply image processing techniques to analyze diseases of citrus fruits with corresponding solutions caused by bacteria, and give information about diseases related to citrus fruits and how to cure them. The researchers used the Spiral Model as a Software Development Life Cycle (SDLC) model to develop this application. In this model, researchers can plan the flow of the application. If the application did not meet the desired result, the researchers could revise it again until it met the desired one. The researchers used the convolutional neural network to classify and process the captured images of the citrus fruits’ diseases and pests. The researchers asked the selected Citrus farmers in Oriental Mindoro to evaluate the project using the different ISO 25010 criteria and rated the application as very acceptable overall.

Detection and Classification of Fabric Defects Using Deep Learning Algorithms

The textile industry primarily relies on fabric as a crucial raw material, the production of which involves multiple complex stages. Due to the multitude and complexity of these stages, fabric defects can frequently occur. With the modern fabric production process being nearly fully automated, and given the variety of potential defects, detecting errors on fabrics has become increasingly challenging. The rapid pace of production and the substantial market share of the sector mean that relying on human inspection for error detection can lead to significant time losses and can reduce the accuracy of defect detection to around 60%. Consequently, recent years have seen a shift towards the development of intelligent systems for fabric defect detection in parallel with technological advancements. With the rapid progression of artificial intelligence, the application of image processing techniques has commenced in this field. This study has developed a real-time defect detection system for fabrics using deep learning techniques. Initially, a network model was created using an open-source neural network library, CNN, achieving 89% accuracy. Subsequent implementations using the VGG16 and InceptionV3 architectures reached accuracies of 89% and 86%, respectively. To further improve the study, fabrics were classified into two categories: defective and non-defective, and the pre-trained Convolutional Neural Networks model ResNet50-v2 was employed as a feature extractor. This approach yielded an approximate accuracy of 95%.

What can machine vision do for lymphatic histopathology image analysis: a comprehensive review

Over the past 10 years, machine vision (MV) algorithms for image analysis have been developing rapidly with computing power. At the same time, histopathological slices can be stored as digital images. Therefore, MV algorithms can provide diagnostic references to doctors. In particular, the continuous improvement of deep learning algorithms has further improved the accuracy of MV in disease detection and diagnosis. This paper reviews the application of image processing techniques based on MV in lymphoma histopathological images in recent years, including segmentation, classification and detection. Finally, the current methods are analyzed, some potential methods are proposed, and further prospects are made.

Enhancing sustainability for pavement maintenance decision-making through image processing-based distress detection

Road maintenance sustainability entails implementing practices that reduce the impact of road accident. while simultaneously ensuring the durability and functionality of the road infrastructure. Pavement distress is a major concern for transportation agencies as it affects the safety and comfort of road users. This paper presents a novel approach for prioritizing pavement distress through the application of image processing techniques and the Analytic Hierarchy Process (AHP). The proposed method involves capturing images of the pavement surface using a high-resolution camera and analyzing them using image processing algorithms. The images are processed to identify different types of pavement distress such as cracks, potholes, and rutting. The severity of each type of distress is then quantified. AHP is utilized to prioritize the identified distress based on its severity and impact on road users. The proposed approach has been tested on real-world pavement images, and the results demonstrate its effectiveness in accurately identifying and prioritizing pavement distress. This method can assist transportation agencies in making informed decisions about maintenance and repair activities, leading to improved road safety and reduced maintenance costs. This automated method achieved an accuracy percentage of about 95%.

A New Model for Predicting Drag Coefficient and Settling Velocity of Coarse Mineral Particles in Newtonian Fluid

Efficient transport in vertical pipeline hydraulic lifting systems is vital for coarse-grained ore, necessitating a deep comprehension of the settling traits of coarse mineral particles. In this study, we conducted a series of settling experiments on individual coarse particles in Newtonian fluids with varying viscosities, employing a self-designed and manufactured settling apparatus. A total of 133 sets of experimental data on the free settling of coarse particles in Newtonian fluids were obtained by recording the particle settling process with a high-speed camera and applying image processing techniques. A mechanical model was employed to perform statistical analysis on the experimental data and establish a predictive model for the drag coefficient and an explicit predictive model for the settling terminal velocity of coarse-grained ore in Newtonian fluids. The average relative errors between the predicted values and experimental values of the drag coefficient and settling terminal velocity models are 4.26% and 7.34%, respectively. This confirms the reliability of the provided predicted model, providing a theoretical foundation for determining the hydraulic lifting speed of coarse mineral particles in vertical pipelines for deep mining.

Application of Image Processing Techniques to Round Corners of Topology-Optimized Motor Rotors Shapes

Application of Image Processing Techniques to Round Corners of Topology-Optimized Motor Rotors Shapes

ScrollTimes: Tracing the Provenance of Paintings as a Window Into History.

The study of cultural artifact provenance, tracing ownership and preservation, holds significant importance in archaeology and art history. Modern technology has advanced this field, yet challenges persist, including recognizing evidence from diverse sources, integrating sociocultural context, and enhancing interactive automation for comprehensive provenance analysis. In collaboration with art historians, we examined the handscroll, a traditional Chinese painting form that provides a rich source of historical data and a unique opportunity to explore history through cultural artifacts. We present a three-tiered methodology encompassing artifact, contextual, and provenance levels, designed to create a "Biography" for handscroll. Our approach incorporates the application of image processing techniques and language models to extract, validate, and augment elements within handscroll using various cultural heritage databases. To facilitate efficient analysis of non-contiguous extracted elements, we have developed a distinctive layout. Additionally, we introduce ScrollTimes, a visual analysis system tailored to support the three-tiered analysis of handscroll, allowing art historians to interactively create biographies tailored to their interests. Validated through case studies and expert interviews, our approach offers a window into history, fostering a holistic understanding of handscroll provenance and historical significance.

A deep convolutional neural network architecture for breast mass classification using mammogram images

<p>Breast cancer is one of the second most common cancer occurring worldwide. Early identification of the disease is a major interest that promises to propose several diagnostic procedures to prevent further surgical interventions. This research paper aims to develop a breast mass classifier system using deep learning to differentiate breast mass images from normal mammographic images. The benchmark mammographic datasets CBIS-DDSM, INbreast, and mini-MIAS are used for constructing the proposed model DELU-BM-CNN. The region of interest is identified by applying image processing techniques (median filter, binarization and dilation) and the images are enhanced and sharpened using adaptive histogram equalization and unsharp masking techniques. The pre-processed images are trained with a minimum of five deep convolutional layers activated by an Exponential Linear Unit (ELU) which is developed from scratch for feature learning and classifying the given whole mammographic images. Dropout, Data normalization, and Global average pooling are some of the regularization techniques adopted to prevent the model from over-fitting. The proposed models are able to classify CBIS-DDSM images with an accuracy of 96.60%, INbreast images with 96.20% and MIAS images with 97.40%. The experimental results are also compared with conventional Rectified Linear Unit (ReLU) and Leaky ReLU activation function that promises the proposed model as a good prognosticator than the state-of-art models for cancer diagnosis using mammogram images as input.</p>

Ice floe segmentation and floe size distribution in airborne and high-resolution optical satellite images: towards an automated labelling deep learning approach

Abstract. Floe size distribution (FSD) has become a parameter of great interest in observations of sea ice because of its importance in affecting climate change, marine ecosystems, and human activities in the polar ocean. A most effective way to monitor FSD in the ice-covered regions is to apply image processing techniques to airborne and satellite remote sensing data, where the segmentation of individual ice floes is a challenge in obtaining FSD from remotely sensed images. In this study, we adopt a deep learning (DL) semantic segmentation network to fast and adaptive implement the task of ice floe instance segmentation. In order to alleviate the costly and time-consuming data annotation problem of model training, classical image processing technique is applied to automatically label ice floes in local-scale marginal ice zone (MIZ). Several state-of-the-art (SoA) semantic segmentation models are then trained on the labelled MIZ dataset and further applied to additional large-scale optical Sentinel-2 images to evaluate their performance in floe instance segmentation and to determine the best model. A post-processing algorithm is also proposed in our work to refine the segmentation. Our approach has been applied to both airborne and high-resolution optical (HRO) satellite images to derive acceptable FSDs at local and global scales.