Digital Image Processing Techniques Research Articles

Deep learning enhanced Watershed for microstructural analysis using a boundary class semantic segmentation

The mechanical properties of the materials are determined by the size and morphology of fine microscopic features. Quantitative microstructural analysis is a key factor to establish the correlation between the mechanical properties and the thermomechanical treatment under which material condition has been achieved. As such, microstructural analysis is a very important and complex task within the manufacturing sector. Published standards are used for metallographic analysis but typically involve extensive manual interpretation of grain boundaries, resulting in measurements that are slow to produce, difficult to repeat and highly subjective. Computer vision and the evolution of artificial intelligence in the past decade can offer solutions to such problems. Deep learning and digital image processing techniques allow digital microstructural analysis to be automated using a fast and repeatable method. This paper proposes a novel boundary class semantic segmentation approach (BCSS) to identify each phase of the microstructure and additionally estimate the location of the grain boundaries. The BCSS is then combined with more traditional segmentation techniques based on the Watershed Transform to improve the identification and measurement of each feature within the microstructure using a new, hybrid automated digital microstructure analysis approach. The new method is validated on a published dataset of two-phase titanium alloy microstructure pictures captured using a scanning electron microscope. Measurements match the level of accuracy of accepted manual standards, and the method is demonstrated to be more reliable than other automated approaches. The influence of the subjective nature of manual labelling, required to train the proposed network, is also evaluated.

Application of Image Processing in Different Machining Processes: A Short and Technical Review

This article discusses the use of digital image processing in a variety of machining processes and the benefits that it brings to the industry. In this article, we will also cover the benefits and drawbacks of using digital image processing techniques instead of the various different sensors that are utilized in machining in order to increase product quality. This article provides a concise introduction to several image processing methods that are utilized in the machining process. This paper contains a discussion of a comprehensive analysis of the applications of image processing that have been used in machining during the past ten years. In addition, an illustration of one approach to image texture analysis that may be applied for cutting tool condition identification through the examination of photographs of machined surfaces is shown. A general conclusion that can be drawn from this and leads to the necessary further research in this area has also been discussed.

Development of a Model for Recognizing Cracks on Concrete Surfaces Using Digital Image Processing Techniques

Development of a Model for Recognizing Cracks on Concrete Surfaces Using Digital Image Processing Techniques

Utilizing Multi-Class Support Vector Machine to Detect Plant Leaf Diseases

Abstract: Plant disease has a large influence on agricultural output and food security. For successful illness management, early detection and correct diagnosis of these disorders are critical. Digital image processing techniques have emerged as a viable tool for automated plant leaf disease identification in recent years. The goal of this study is to use image processing to create a more efficient and accurate system for detecting leaf illness. K-means clustering, Gray-Level Co-Occurrence Matrix (GLCM), and Support Vector Machine are used to create the model. Kmeans clustering is employed for image segmentation, GLCM, and multiclass SVM for feature extraction and illness classification respectively. The maximum accuracy is 98.4%

Quantitative detection of internal defects in objects using holographic double exposure

Recent years have witnessed an increasing demand for high-quality materials in all industries; therefore, internal defects are disastrous for maintaining the required industrial quality. In this study, qualitative and quantitative analysis of the internal defects of objects is conducted using differential holographic double-exposure and digital image processing techniques. The objects underwent thermal loading, and two holograms were recorded at different temperatures. Internal defects were qualitatively and quantitatively analysed by processing and comparing the interference fringes produced by removing and reproducing the two holograms via digital image processing. The fringes at the location of the sample defects differed from the changes in surrounding fringes, and the defect size was determined by fitting the fringe region at the defect location. Finally, the experimental results were validated using a structured light 3D measurement method to identify the size of the defects. Interesting results were obtained, with a maximum error of 9.637 % for irregular defects and a minimum error of 3.528 % for regular defects.

DETEKSI AREA KERUSAKAN PADA CITRA TERUMBU KARANG AKIBAT CORAL BLEACHING BERBASIS PENGOLAHAN CITRA DIGITAL

Coral reefs have a very large role in marine ecosystems, because they provide rich habitats and support biodiversity. Segmentation on images of coral reefs experiencing coral bleaching is also very important. Coral bleaching is a phenomenon in which coral reefs lose their symbiotic algae pigments, resulting in bleaching and possibly death of coral reefs. By segmenting coral reef images that experience coral bleaching, we can separate areas affected by bleaching from areas that are still healthy. This study aims to detect areas of damage that occur in underwater images of coral reefs that have experienced coral bleaching. By using techniques in digital image processing and segmentation based on colour intensity, satisfactory results are obtained. The test results show that the RAE and ME values are quite low, namely 0.051 and 0.035 respectively with an average processing time of 0.2 seconds. This research is expected to assist in further analysis and modelling related to coral bleaching to understand the causative factors and develop appropriate protection strategies.

Integrated digital image processing techniques and deep learning approaches for wheat stripe rust disease detection and grading

This study proposes integrated image processing and deep learning approaches for wheat rust disease detection and grading. Wheat stripe rust is one of the diseases affecting the crop in nearly all wheat-growing countries, where humidity and high temperatures characterize the growing season. The early and prompt detection of this disease is essential for establishing timely remedial measures to contain its spread. We present several image-processing and deep-learning approaches for detecting wheat stripe rust. The image processing techniques used in this research include single-band and dual-band processing, computing the Visual Atmospheric Resistance Index (VARI), image segmentation, and image cropping. These techniques are evaluated with different deep learning approaches like transfer learning and other Convolutional Neural Networks (CNNs) architectures. The results obtained on the dataset show that processing the images to eliminate the background as much as possible before training a suitable deep learning model with a low number of epochs optimally classifies the wheat stripe rust into the target classes (i.e., healthy, resistant, and susceptible) with an overall accuracy of 84.1% on the holdout test set. Besides the comparative analysis of various image processing approaches with different deep learning models, this study suggests multiple future research directions.

Numerical study of the gas–solid flow pattern in methanol to olefiens fluidized bed reactor: Cold model

AbstractIn the present study, the hydrodynamic behaviour of a three‐dimensional bubbling fluidized bed reactor was studied using the finite volume method and granular kinetic theory. The effect of two types of gas distributors (i.e., perforated and porous plates) on the gas/solid hydrodynamic was investigated. The residence time distribution model was utilized to check the deviation of the gas flow pattern from the ideal flow patterns for two types of distributors. The parameters indicating the degree of back‐mixing, such as mean residence time and dimensionless variance, were calculated. Also, the model was verified through the comparison of the particle velocity with experimental data obtained by the particle image velocimetry–digital image processing (PIV‐DIA) technique. It was found that the gas flow pattern in the present bubbling fluidized bed is closer to the mixed flow than the plug flow pattern. It was also found that the porous plate distributor gives more uniform hydrodynamic distributions of gas and solid in the bed, which results in better gas–solid mixing due to the formation of the smaller bubbles.

Microstructural evolution and coarsening behavior of the precipitates in 2205 duplex stainless steel aged at 850 °C

The formation of the secondary phases in the 2205 duplex stainless steel (DSS) influences greatly its mechanical properties. Studies on the microstructural evolution and coarsening behavior of the precipitates in the 2205 DSS have both scientific and technological significance. In the present work, the evolution of the composition and morphology of the precipitates in the 2205 DSS aged at 850 °C for periods up to 200 h are systematically investigated by SEM/EDS (Scanning Electron Microscope/Energy Dispersive Spectrometer) and TEM (Transmission Electron Microscope). In addition, the size statistics for intermetallic precipitates, especially for the σ phase, are carried out based on the SEM images by using deep learning and digital image processing technique. Based on the above, the average interphase energy between the σ and γ phases is reasonably estimated according to the Ostwald coarsening mechanism. This work provides a comprehensive understanding of the microstructural evolution and coarsening behavior of the precipitates in 2205 DSS.

Assessing the Photovoltaic Power Generation Potential of Highway Slopes

The solar photovoltaic (PV) power generation system (PGS) is a viable alternative to fossil fuels for the provision of power for infrastructure and vehicles, reducing greenhouse gas emissions and enhancing the sustainability of road transport systems. A highway slope is generally an idle public area with high accessibility, which is the ideal application scenario for a PV PGS. The assessment of PV power generation potential (PGP) is key for the planning and design of PV PGS projects. Previous approaches to potential assessments are mainly based on digital maps and image processing techniques, which do not fully consider the impacts of the highway orientation, the slope geometric characteristics, and the PV panel placement scheme on the evaluation results. Therefore, this study proposes an assessment method for the PV PGP on highway slopes using the design or calculated highway and slope geometric parameters and the solar radiation received by PV panels under the desirable placement scheme. Highway segmentation and geometric parameter calculation methods were established, and the optimal PV array placement schemes for typical slope orientations were determined by simulating the PV power generation in the software PVsyst (version 7.2). Afterwards, the theoretical PGP could be calculated using the received solar radiation and the available slope area. By subtracting the energy loss caused by temperature changes, the operation of inverters, and the PV modules’ performance decay, the actual PV PGP could be obtained. Finally, a case study of the solar PGP assessment of a 1.97 km long highway section is provided, and the feasibility of the proposed method is verified.

An Automatic Approach for Bone Tumor Detection from Non-Standard CT Images

Image processing techniques are applied in many fields of science. This study aims to detect tumors in the foot and create 3D models via computed tomography (CT), as well as to produce biometric data. 1 039 CT images were obtained from a server. The parameters used were a collimation of 64 detectors, a scanning thickness of 0,5-3 mm, and a pixel size of 512 x 512, with a radiometric resolution of the 16-bit gray levels. Noise reduction, segmentation, and morphological analysis were performed on CT scans to detect bone tumors. In addition, this study used digital image processing techniques to create a virtual three-dimensional (3D) model of bone tumors. The performance of our proposal was evaluated by analyzing the receptor operating characteristics (ROC). According to the results, the sensitivity, specificity, and precision in tumor detection were 0,96, 1, and 0,98%, respectively, with a 0,99% average F-measure. Radiologist reports were used for the sake of comparison. The proposed technique for detecting bone tumors of the foot via CT can help radiologists with its increased precision, sensitivity, specificity, and F-measure. This method could improve the diagnosis of foot and ankle tumors by allowing for the multidirectional quantification of abnormalities.

Ultraviolet Radiation Transmission in Building’s Fenestration: Part II, Exploring Digital Imaging, UV Photography, Image Processing, and Computer Vision Techniques

The growing demand for sustainable and energy-efficient buildings has highlighted the need for reliable and accurate methods to detect fenestration deterioration and assess UV radiation transmission. Traditional detection techniques, such as spectrophotometers and radiometers, discussed in Part I, are often expensive and invasive, necessitating more accessible and cost-effective solutions. This study, which is Part II, provides an in-depth exploration of the concepts and methodologies underlying UV bandpass-filtered imaging, advanced image processing techniques, and the mechanisms of pixel transformation equations. The aim is to lay the groundwork for a unified approach to detecting ultraviolet (UV) radiation transmission in fenestration glazing. By exploiting the capabilities of digital imaging devices, including widely accessible smartphones, and integrating them with robust segmentation techniques and mathematical transformations, this research paves the way for an innovative and potentially democratized approach to UV detection in fenestration glazing. However, further research is required to optimize and tailor the detection methods and approaches using digital imaging, UV photography, image processing, and computer vision for specific applications in the fenestration industry and detecting UV transmission. The complex interplay of various physical phenomena related to UV radiation, digital imaging, and the unique characteristics of fenestration glazing necessitates the development of a cohesive framework that synergizes these techniques while addressing these intricacies. While extensively reviewing existing techniques, this paper highlights these challenges and sets the direction for future research in the UV imaging domain.

Improving Endoscopic Image Quality through the Use of High Dynamic Range Imaging-Like Method with Real-Time Performance.

High Dynamic Range (HDR) imaging is a digital image processing technique used to produce a wider range of brightness and color by using multiple captures of a scene taken with different exposures times. It enables capturing more details and producing a more natural-looking image with less washed-out highlights and deeper, more saturated colors. In medical endoscopy, HDR imaging enhances the visibility and clarity of images captured during endoscopy procedures. It provides enhanced visualization of subtler details in both dark cavities and bright areas, resulting in a uniformly exposed view and improved contrast among various tissue types. Standard HDR imaging methods are often complex and computationally demanding, making them unsuitable for performance-critical applications like endoscopy, where real-time performance is crucial. This paper introduces a more efficient and less complex method for achieving HDR-like image quality in real time. The method takes a high-pixel-bit-depth frame and generates multiple low-pixel-bit-depth frames and uses them to generate the high quality image. The focus of the paper is to enhance endoscopic image quality using HDR imaging, and the proposed method is demonstrated to be effective in achieving this goal with real-time performance. The method is implemented in the FPGA System-on-a-Chip (SoC) of a bronchoscope video processor system, and its effectiveness is verified through a simulated study using a phantom, which confirms the improved image quality and real-time performance.

Spatial-temporal mapping of forest vegetation cover changes along highways in Brunei using deep learning techniques and Sentinel-2 images

Infrastructure development is a leading driver of forest cover loss in the tropics, resulting in a significant decrease in biodiversity. With recent advancements in digital image processing and satellite imagery techniques, this study presents an approach for tracking forest cover alterations along a recently constructed highway in Brunei. Specifically, this paper harnesses the power of cutting-edge deep learning (DL) models; U-Net and Deeplabv3+, to classify forest and non-forest features from Sentinel-2 satellite images captured along the Telisai–Lumut highway between 2018 and 2022. A computational performance analysis and network comparison revealed that the Deeplabv3+ had a remarkable semantic segmentation performance, with a mean intersection over union of 0.9247, followed by an ensemble of four networks, with a mean intersection over union of 0.9171. The best model consistently outperformed other models in segmenting forest and non-forest features, with an intersection over union (IOU) of 0.9463 and 0.9031, respectively. Moreover, deploying hyperparameter optimization allowed us to assess the model's sensitivity to different hyperparameters for better results. Furthermore, forest change statistics were computed using the best model's generated forest change maps, and the results indicated a 16.7% gain in the overall forest vegetation cover along the highway, demonstrating the validity of the results. These reproducible findings will provide forest managers and ecologists with near real-time information on the status, health, and extent of forest ecosystems. This study's results have critical implications for using DL to monitor forest cover changes and will be of great interest to the scientific community as they open new avenues for future research and underline the need for continued innovation in this important field.

Detection of Early Blight Tomato Leaf Using k-Means Clustering

Early blight is one of the major diseases of tomatoes that affects the leaves and fruit quality. Detection and estimation of the disease severity are performed using the visual observation method. Visual detection requires significant time for visual inspection of a large cultivated area. Thus, image processing techniques have proven to be an effective method as compared to visual analysis. In this study, digital image processing methods and techniques were used to detect early blight of tomato (EBT), estimate the disease severity, and classify tomato leaves. Totally, 198 infected plants were randomly taken from the Haramaya University research site "Rare" at four different times. Diseased potato leaf images were captured, resized, and stored for experimentation. The stored images were processed using median filtering to remove noise while preserving useful features in an image and image enhancement. The RGB images were transformed to gray scale and CIELAB color space, and the k-means clustering was applied to estimate the disease severity of the potato leaves, and Otsu’s thresholding algorithm was applied to estimate the disease severity of both the detached and live leaves. MATLAB algorithms will be developed to determine the total area and infected lesion area of the leaf samples.

Motor unit placement in a realistic muscle cross section model: performance of a new algorithm and effects of muscle architecture on surface EMG power spectral components

Objetive. Surface electromyogram (sEMG) characteristics are strongly dependent on the spatial localization of motor units (MU). Thus, the definition of the muscle architecture is a crucial step in the simulation of sEMG. Muscles cross section or even the shape of the MU territories is not a regular geometric figure. Moreover, MU may have different fibre densities, and its distribution may not be random but follow a spatial regionalization by type. Despite these prior conditions, the method for MU placement in a muscle model should be able to achieve constant muscle fibre density, which is a generally accepted property of any muscle. Approach. An algorithm was developed that meets the previously established requirements for the distribution of MU in a muscle model. The algorithm considers the muscle cross section and MU territories as bitmaps and encodes the muscle fibre density distribution by pixel brightness. This allowed the use of digital image processing techniques to optimally distribute the MU. The impact of the different muscle architectures on sEMG was investigated using one of the most accepted models of motor unit pool recruitment combined with a model of MU action potential simulation. element-citationitalic Main results. The algorithm produced a completely regionalized architecture with constant muscle fibre density in a muscle with both an irregular cross-section and elliptic MU territory. Significance. In addition to the improvement in the realism of the muscle simulations, the algorithm allowed the investigation of the influence of the spatial distribution of MU on the sEMG power spectrum, thus helping to explain the existing inability to detect changes in motor unit recruitment strategies through the spectral analysis of interference patterns.

Remote sensing and GIS tool application for tracing the water body of Bangalore area

To comprehend the function of the water bodies at various sizes, such as regional and global, for example, in biogeochemical cycles, it is necessary to more correctly estimate the size, form, and extent of the water bodies. Excluding the few areas with precise maps, scaling upward to vast area regions relies on statistical estimates of the number and size of lakes, rivers, and other water bodies, which explains why estimates are wrong. In comparison to the currently available conventional and frequently used methods of mapping land surface water (LSW), remote sensing offers several extra advantages since it is a more accurate, efficient, and reliable information source with the ability to make multiple high-frequency and repeatable quantities of observations. This project describes the development and technological evolution of the Water Body Map (WBM), which employs an automated algorithm to interpret multi-temporal Landsat images from the USGS database. In this project, we used Landsat 5 band photographs to define a water body map without any cloud cover. Permanent water bodies were separated from temporary water-covered areas by estimating the frequency of water bodies in overlapping, multi-temporal Landsat images. In contrast to past studies, the WBM now makes a more precise distinction between rivers and floodplains by looking at the frequency of water bodies. As a result, it is clear that using multi-temporal images from sources is equally as important for mapping global water bodies as analysis at a higher resolution. The classification's accuracy was confirmed in Hokkaido (Japan) and the United States of America using databases of actual water bodies. Around 70% of lakes N1 km2 have relative water area errors of b25%, and the commission error was diminishing and nearly nonexistent. The Water Body Mapping's overall accuracy should be helpful for much larger scale research projects in hydrology, biogeochemistry, and climate systems, among other things. It can also include a quantification of the kinds of water bodies that are currently present around. Some smaller water bodies were overlooked, primarily because shoreline pixels were not included. The primary goal of this project was to estimate the accuracy of mapping water bodies using Landsat 5 TM data by employing straightforward digital image processing techniques.

Authorized Automatic Vehicle Allowance System

Abstract: In most organizations, the ongoing vehicle entry registration process for visitors, staff, or students entering the organization includes a security guard checking the driver's identification card or looking for a membership sticker on the vehicle's windscreen, which can easily be lost or stolen. This writing process is tiresome and time-demanding, and it is prone to incorrect notes; additionally, backup and sharing of this vehicle information is difficult due to the data being a physical copy. The goal of the Automatic Authorized Vehicle Allowance System is to create a digital picture processing system with the Raspberry Pi that can be used in an entrance parking system. The Raspberry Pi will identify the car with an ultrasonic sensor and photograph it. The device then processes the picture using digital image processing techniques. The system incorporates various algorithms such as localization, normalization, orientation, segmentation, and optical character recognition (OCR), which examines picture characteristics and forecasts license plate numbers. The obtained number plate information is then cross-checked with records stored in a pre-existing database. An administration is in charge of designating a "whitelist," which is a list of vehicles that can join the suggested system. If the car registration is found in the database, a servo motor circuit connected to a Raspberry Pi will open the gate, and after a monitor delay, the gate will shut. The system is built in such a way that it is simple to apply in a short period of time and is cost-effective.

Application of Digital Image Processing Techniques to Detect Through-Thickness Crack in Hole Expansion Test

Advanced high-strength steels (AHSS) have become increasingly popular in the automotive industry due to their high yield and ultimate tensile strengths, enabling the production of lighter car body structures while meeting safety standards. However, they have some setbacks compared to conventional steels, such as edge cracking through sheet thickness caused by forming components with shear-cut edges. When characterizing the formability of sheet metal materials, the hole expansion test is an industry-standard method used to evaluate the stretch-flangeability of their edges. However, accurately visualizing the first cracking is usually tricky and may be subjective, often leading to inconsistent results and low reproducibility with some impact of the operator on both direct and post-processing measurements. To address these issues, a novel digital image processing method is presented to reduce operator reliance and enhance the accuracy and efficiency of the hole expansion test results. By leveraging advanced image processing algorithms, the proposed approach detects the appearance of the first edge cracks, enabling a more precise determination of the hole expansion ratio (HER). Furthermore, it provides valuable insights into the evolution of the hole diameter, allowing for a comprehensive understanding of the material behavior during the test. The proposed method was evaluated for different materials, and the corresponding HER values were compared with the traditional method.

Robust deep learning-based fault detection of planetary gearbox using enhanced health data map under domain shift problem

Abstract The conventional deep learning-based fault diagnosis approach faces challenges under the domain shift problem, where the model encounters different working conditions from the ones it was trained on. This challenge is particularly pronounced in the diagnosis of planetary gearboxes due to the complicated vibrations they generate, which can vary significantly based on the system characteristics of the gearbox. To solve this challenge, this paper proposes a robust deep learning-based fault-detection approach for planetary gearboxes by utilizing an enhanced health data map (HDMap). Although there is an HDMap method that visually expresses the vibration signal of the planetary gearbox according to the gear meshing position, it is greatly influenced by machine operating conditions. In this study, domain-specific features from the HDMap are further removed, while the fault-related features are enhanced. Autoencoder-based residual analysis and digital image-processing techniques are employed to address the domain-shift problem. The performance of the proposed method was validated under significant domain-shift problem conditions, as demonstrated by studying two gearbox test rigs with different configurations operated under stationary and non-stationary operating conditions. Validation accuracy was measured in all 12 possible domain-shift scenarios. The proposed method achieved robust fault detection accuracy, outperforming prior methods in most cases.