Image Processing Technology Research Articles

Flow behavior and size characteristics of microbubble swarm in a HiGee-aided fixed bed reactor with different packed structures

A HiGee-aided fixed bed reactor (HFBR) has great potential for catalytic hydrogenation with hydrogen microbubble swarm for gas–liquid mass transfer intensification. However, the flow behavior and size characteristics of microbubble swarm in the catalyst bed of HFBR still remain unclear. In this work, the flow characteristics of microbubble swarm in different packed structures of catalysts and catalyst supports had been fully analyzed by the high-speed photography and image processing technology. The optimal ratio of packing heights for the bottom of the catalyst supports was Φ13 mm: Φ6 mm: Φ3 mm = 3: 1: 1, while the top of the catalyst supports was packed with single catalyst support of Φ6 mm. Furthermore, a dimensionless correlation based on the experimental results had been developed to predict (d32)o/(d32)i with deviations within ±15 %. The effective interfacial area of HFBR was 0.6 ∼ 6.1 × 103 m2/m3 under the experimental conditions.

Hydraulic fracturing simulation of concrete dam integrating intelligent crack detection and refined modeling methods

Concrete dams are prone to cracks during construction and operation process. Cracks on the upstream surface may induce the risk of hydraulic fracturing, thereby significantly reducing the safety of dam structure. Nevertheless, image processing technology, non-destructive testing technology and structural performance analysis technology with cracks are in a relatively independent state, leading to difficulties in analysis of actual cracks combined with numerical simulations. This paper proposes a novel approach for combination of intelligent identification technology and refined finite element method (FEM) for modeling and simulating the hydraulic fracturing resistance of the actual crack in concrete dams. With image processing technology and non-destructive testing technology, direction and depth of actual cracks are identified and extracted, then the 3D geometric model of actual cracks is generated. The refined model with the identified actual crack is constructed by the node projection strategy. This approach is used to assess the hydraulic fracturing resistance of the concrete dam with actual cracks in the alpine region. The results demonstrate that the critical crack depth of hydraulic fracturing is 12.93 m, and the opening width is 1.01 mm, which is 2.53 times the limit value for Class D cracks in the code (DL/T5251–2010). The refined modelling method based on actual crack detected on site can accurately evaluate the crack state compared to the conservative code.

Deep-learning based in-situ micrograph analysis of high-density crystallization slurry using image and data enhancement strategy

Process analytical technology (PAT) plays a crucial role in optimizing crystalline powder product qualities, improving process stability, and reducing experimental costs by enabling real-time monitoring and control of process variables during the crystallization process. In-situ process imaging and analysis have gained significant attention due to their capacity to provide abundant information through images, leading to remarkable advancements in image processing technology, particularly with the support of artificial intelligence. However, the performance suffers when processing high-density slurry images due to challenges such as defocusing, overlapping, background and crystal edges blurriness, and inconsistencies in crystal scales. To address these challenges and explore research strategies based on deep learning for segmenting and analyzing high slurry density images effectively, this study constructed a specific dataset that contains high-density slurry crystal images which is well-labeled, and introduced state-of-the-art neural networks including YOLOv8, U2-net, and Mask R-CNN combined with image and data enhancement strategies. Comparative analysis revealed that YOLOv8 outperformed Mask R-CNN and U2-net by capturing multi-dimensional information in high-density crystal slurry scenarios. Finally, based on the segmentation results of the practical taurine crystallization process using proposed strategy, 1D crystal size, aspect ratio along with 2D size distributions were extracted for further evaluation of crystallization kinetics and crystal properties accurately and quickly.

Experimental verification and rapid estimation of uncalibrated cable force via video-based and vibration-based measurements.

The stayed-cable is an important component of cable-stayed bridges, with cable force being a focal point during construction and bridge operation. The advancement of camera and image processing technology has facilitated the integration of computer vision technology in structural inspection and monitoring. This paper focuses on enhancing cable force measurement methods and addressing the limitations of traditional testing techniques by conducting experimental research on cable force estimation using video recording. The proposed approach involves capturing video footage of the target on the cable with a smartphone. Subsequently, a combination of techniques such as the background subtraction method, image morphology processing, and Hough transform image processing technology are employed to detect the precise center coordinates and ultimately obtain the accurate displacement-time curve of the cable's vibration. In addition, the graphic Circularity Coefficient (CC) has been introduced to assess its effectiveness in post-motion-blur image processing for circular targets. The fundamental frequency of the cable is determined by the fast Fourier transformation, and the relationship between the cable force and the fundamental frequency is used to estimate the cable force. The experimental results are compared with data from accelerometers and force gauges, demonstrating that the frequency measurement error is below 1.2% and the cable force test error is less than 3%. In the process of acquiring the cable's fundamental frequency, the test directly employs the pixel as the displacement unit, eliminating the need for image calibration. The innovative use of the CC in processing motion-blurred targets ensured accurate recognition of target coordinates. The experimental findings highlight the method's simplicity, speed, and accuracy.

Virtual Reality Surgical Simulations Using Fusion Three-Dimensional Images

In this section, we defined virtual reality(VR)surgical simulations using fusion three-dimensional(3D)images, which are 3D images created by fusing multiple medical image data. The more detailed the fusion of 3D images, the more knowledge and effort are required. In addition, 3D fusion images vary greatly with each case and depend on the skill and orientation of the image creator and the image processing software used. Some creators produce a fused 3D image with ample details to simulate tissue deformation, whereas others are limited to rough observations and use two-dimensional cross-sectional images for detailed anatomical information. Thus, there is no gold standard for creating fused 3D images or VR surgical simulations. Therefore, it is important to clarify the objective of a VR surgical simulation. An understanding of image-processing technology is useful in terms of software selection and image-processing efficiency. This section outlines the construction of fused 3D images and the use of VR surgical simulations based on actual clinical applications.

Line segment detection algorithm in image extraction improvement study

In recent years, image processing technology has been developing and maturing, but due to the influence of many interfering factors in the acquisition process, there is a large amount of redundant information in the images obtained. The line segment detection algorithm in image extraction needs to be improved. This study utilizes computer technology to improve the line segment detection technology, and designs a line segment detection algorithm based on the linear detection improvement. Firstly, based on the basic principle of straight line detection algorithm, for the problems of line segment breakage and missing in straight line detection, RGB three-channel grayscale map is applied to detect line segments. Then the detected line segments are connected, merged and deleted. The test results show that the line segment detection algorithm improved based on straight line detection has the highest accuracy rate of 94.50 %, and the average processing time per image is also the lowest at 0.2 s. The algorithm runs faster at 0.25 s and has a higher F-value. It is able to detect the boundaries of a variety of rectangular targets, using the improved line segment detection algorithm has a wide range of applicability, lower error rate, and strong anti-interference ability. The improved line segment detection algorithm has a greater advantage in rectangular target extraction for document, text and book type images.

Automated Physiological Status Detection and Disease Evaluation of Critically Ill Patients via Image Processing Technologies

Automated Physiological Status Detection and Disease Evaluation of Critically Ill Patients via Image Processing Technologies

Coffee disease detection and classification using image processing: A Literature review

Coffee, as one of the world's most consumed beverages, sustains livelihoods for millions across more than 50 nations. The vulnerability of coffee plants to diseases, particularly Coffee Leaf Rust and Coffee Berry Disease, poses a significant threat to global production and quality. Leveraging advancements in image processing and computer vision, researchers have explored diverse classification algorithms, ranging from traditional Support Vector Machines to state-of-the-art Deep Convolutional Neural Networks (DCNNs). This review literature addresses the challenges of coffee disease detection, emphasizing the need for precise and early identification. Notable studies have achieved commendable accuracies, such as SVMs reaching 96% and DCNNs demonstrating precision but with extended training times. Innovations like feature concatenation, transfer learning, and ensemble methods have emerged as strategies to overcome classification limitations. Recent breakthroughs showcase impressive results, including DenseNet models achieving a classification accuracy of 99.57% and MobileNetV2 reaching 99.93%. Additionally, Convolutional Neural Networks and VGG-19 architecture demonstrated a promising F1-Score of 90% in classifying various coffee leaf diseases. This concludes with a vision for ongoing advancements, emphasizing the fusion of image processing and machine learning technologies to safeguard the global coffee industry by enabling early and accurate disease detection.

Flame‐assisted ultrafast synthesis of functionalized carbon nanosheets for high‐performance sodium storage

AbstractThe unique structural features of hard carbon (HC) make it a promising anode candidate for sodium‐ion batteries (SIB). However, traditional methods of preparing HC require special equipment, long reaction times, and large energy consumption, resulting in low throughputs and efficiency. In our contribution, a novel synthesis method is proposed, involving the formation of HC nanosheets (NS‐CNs) within minutes by creating an anoxic environment through flame combustion and further introducing sulfur and nitrogen sources to achieve heteroatom doping. The effect of heterogeneous element doping on the microstructure of HC is quantitatively analyzed by high‐resolution transmission electron microscopy and image processing technology. Combined with density functional theory calculation, it is verified that the functionalized HC exhibits stronger Na+ adsorption ability, electron gain ability, and Na+ migration ability. As a result, NS‐CNs as SIB anodes provide an ultrahigh reversible capacity of 542.7 mAh g−1 at 0.1 A g−1, and excellent rate performance with a reversible capacity of 236.4 mAh g−1 at 2 A g−1 after 1200 cycles. Furthermore, full cell assembled with NS‐CNs as the can present 230 mAh g−1 at 0.5 A g−1 after 150 cycles. Finally, in/ex situ techniques confirm that the excellent sodium storage properties of NS‐CNs are due to the construction of abundant active sites based on the novel synthesis method for realizing the reversible adsorption of Na+. This work provides a novel strategy to develop novel carbons and gives deep insights for the further investigation of facile preparation methods to develop high‐performance carbon anodes for alkali‐ion batteries.

Comparative Analysis of Deep Learning Models for Olive Detection on the Branch

The future of deep learning integration in agriculture holds great potential for advancing sustainable agricultural practices, precision agriculture and improved decision-making. With the rapid development of image processing and artificial intelligence technologies in recent years, deep learning has begun to play a major role in identifying agricultural pests and optimizing agricultural product marketing. However, there are challenges related to data quality, model scalability, and geographical limitations for widespread adoption of deep learning in agriculture. This study on Olive was conducted to improve the quality of the data set and to ensure more reliable training of object detection models. According to the result of the training process of YOLOv7 used in the study, it was concluded that it was characterized by decreasing loss values and showed an increase in the model's ability to detect objects correctly. It was observed that the other model, YOLOv8l, had a more effective learning capacity and a tendency to learn faster. The performance of both models was evaluated with various metrics, and it was determined that YOLOv8l had higher Precision, Recall, and mAP values. It was emphasized that YOLOv8l showed high performance even in low epoch numbers and can be preferred especially in cases where time and computational resources were limited. It was determined that YOLOv7 made detections in a wide confidence range, but had difficulty in detections with low confidence scores. It was observed that YOLOv8l made more stable and reliable detections with higher confidence scores. The metric data of the "YOLOv8l" model was found to be higher compared to other models. The F1 score of the YOLOv5l model was 92.337%, precision 96.568%, recall %88,462,mAP@0.5:0.65 value gave the highest score with 94.608%. This research on deep learning-based object detection models indicated that YOLOv8l showed superior performance compared to YOLOv7 and was a more reliable option for agricultural applications.

Deep Learning Based Object Detection with Unmanned Aerial Vehicle Equipped with Embedded System

This study aims to introduce an Unmanned Aerial Vehicle (UAV) platform capable of performing real-time object detection and classification tasks using computer vision techniques in the field of artificial intelligence. Previous scientific research reveals the utilization of two different methods for object detection and classification via UAVs. One of these methods involves transmitting the acquired UAV images to a ground control center for processing, whereafter the processed data is relayed back to the UAV. The other approach entails transferring images over the internet to a cloud system, where image processing is conducted, and the resultant data is subsequently sent back to the UAV. This allows the UAV to autonomously perform predefined tasks. Enabling the UAV with autonomous decision-making capabilities and the ability to perform object detection and classification from recorded images requires an embedded artificial intelligence module. The ability of the UAV to utilize image processing technologies through embedded systems significantly enhances its object detection and classification capabilities, providing it with a significant advantage. This enables the UAV to be used more effectively and reliably in various tasks. In the proposed approach, image processing was achieved by mounting a Raspberry Pi 4 and camera on the UAV. Additionally, a Raspberry Pi-compatible 4G/LTE modem kit was used to provide remote intervention capability, and the Coral Edge TPU auxiliary processor was used to increase object detection speed. The TensorFlow Library and the SSD MobilNetV2 convolutional neural network model were used for image processing. During test flights, accuracy values of approximately 96.3% for car detection and 96.2% for human detection were achieved.

Quantitative analysis for the microstructure of lithium-based grease: Methodology and application

The advancement of image processing technology has introduced novel approaches for material structural analysis. This study develops a quantitative analysis method for the microstructure of lithium-based greases utilizing ImageJ. The SEM images of grease fibers are firstly pre-processed in order to enhance the accuracy of calculation. Then the void fraction in the fiber network is obtained through threshold and area calculation. The average fiber diameter and length are calculated via local thickness analysis and skeleton analysis, respectively. Additionally, this method was applied to assess the microstructural changes in lithium grease under conditions of intense mechanical shearing and high temperature. The possible reasons for the deterioration of grease performance are also discussed.

MCMC algorithm based on Markov random field in image segmentation.

In the realm of digital image applications, image processing technology occupies a pivotal position, with image segmentation serving as a foundational component. As the digital image application domain expands across industries, the conventional segmentation techniques increasingly challenge to cater to modern demands. To address this gap, this paper introduces an MCMC-based image segmentation algorithm based on the Markov Random Field (MRF) model, marking a significant stride in the field. The novelty of this research lies in its method that capitalizes on domain information in pixel space, amplifying the local segmentation precision of image segmentation algorithms. Further innovation is manifested in the development of an adaptive segmentation image denoising algorithm based on MCMC sampling. This algorithm not only elevates image segmentation outcomes, but also proficiently denoises the image. In the experimental results, MRF-MCMC achieves better segmentation performance, with an average segmentation accuracy of 94.26% in Lena images, significantly superior to other common image segmentation algorithms. In addition, the study proposes that the denoising model outperforms other algorithms in peak signal-to-noise ratio and structural similarity in environments with noise standard deviations of 15, 25, and 50. In essence, these experimental findings affirm the efficacy of this study, opening avenues for refining digital image segmentation methodologies.

Implementation of intelligent healthcare image detection system based on OpenCV

As an interface between image processing technology and the real world, image detection can achieve real-time capture of information such as images and movements. It has wide applications in fields such as medical rehabilitation, military simulation, and intelligent recognition. This article aims to achieve autonomous medical detection and assist doctors in diagnosis. Based on a detailed analysis of the principles of image detection technology and the current development status of image detection technology, a comprehensive medical detection system solution is designed. Addressing issues such as the tension of hospital registration queues, lengthy doctor-patient consultations, lack of auxiliary diagnostic functions, and non-visualizable diagnosis and treatment results, this study utilizes literature analysis and modeling to simulate the collection and intelligent analysis of surface images of the human body and limb movements. This enables the realization of functions such as image detection, image capture, image transmission, image processing, and medical feedback. This article has completed the design of an intelligent medical system, solving the problem of low medical diagnosis efficiency. At the same time, it has been found that the clarity of infrared thermal imaging images is not high, which affects the accuracy of collecting human surface images and limb movements. The processing of infrared thermal imaging images has become a direction that needs further research.

Exploring multiple algorithms for leaf classification: A comparative study

In the world there are an estimated nearly half a million plant species, and the leaves of each plant have their own unique characteristics. However, leaf classification has historically been problematic, with some similar-looking leaves being repeatedly identified and errors may even occur. In the past, people used precision instruments, chemical analysis, and other methods to improve the success rate of resolution, but this often required a certain amount of professional knowledge, sometimes it was trouble to operate. And the development of computer in recent years makes the leaf classification problem can be solved better through the image processing and machine learning technology. Therefore, the topic of this paper is to classify leaves by different algorithms and compare the classification results, which include KNN, Random Forest, the neural network algorithm ANN, and CNN. It can be seen through experiments that the classification results obtained by ANN have a higher accuracy rate among these algorithms.

Detection of Rice Leaf Pests Based on Images with Convolution Neural Network in Yollo v8

Detection of rice leaf pests is important in agriculture because it can help farmers determine appropriate preventive measures. One method that can be used to detect rice leaf pests is digital image processing technology. In this research, proof of suitability for solving this case was carried out between the Convolutional Neural Network (CNN) algorithm which was run offline with R-CNN and YOLOv8 for detecting rice leaf pests. At the data preparation stage, images of rice leaves were taken from various sources with a total of 100 images taken from website data and 10 images taken from the research site. Next, preprocessing and data augmentation are carried out to improve image quality and increase data variation. At the model training stage, a training and evaluation process is carried out using two types of algorithms, namely R-CNN and YOLOv8. The accuracy of the testing results using the same data using Yolov8 obtained 87.0% accuracy and 79% precision, while using R-CNN the results obtained were 85% for accuracy and 75% for precision with data divided into 80 training data 20 validation data and 10 testing data. Labeling the dataset uses Makesensei which has been completely standardized, with the resulting parameters being the spots on rice leaves.

Effect of natural fractures on mechanical properties and fracture patterns of shale at microscopic scale: an example from the Lower Cambrian Niutitang formation in Qianbei region

In order to explore the influence of natural fractures on the mechanical properties and failure modes of shale at the micro scale, uniaxial compression numerical experiments were conducted on the shale of the Niutang Formation in northern Guizhou with different natural fracture angles using a rock failure process system and digital image processing technology. It is shown that the compressive strength of shale increases with the increase of natural crack inclination, and the growth rate of shale compressive strength also increases. Shale's microscopic fractures can generally be classified into four categories. The first category is to sprout along the natural cracks to the outside of the shale, and eventually form a crack similar to the "X" type (0°); the second category is to sprout along the natural cracks to the middle and outside of the shale, and eventually form an inverted "Y" type crack (15°, 30°); the third category is to sprout along the natural cracks to the middle and outside of the shale, and eventually form an inverted "Y" type crack (15°, 30°); the second type sprouts along the natural fractures toward the middle and outside of the shale, forming inverted "Y"-type fractures (15°, 30°); the third type cracks along the sides of the natural fractures, forming "Y"-type fractures (45°); and the fourth type does not crack along the natural fractures, forming "S"-type fractures (60°, 75°, and 90°). In the low natural fracture dip shale model, tensile damage mainly occurs, accompanied by a small amount of compressive shear damage; in the high natural fracture dip shale model, tensile damage and compressive shear damage account for a larger proportion in the fracture process.This suggests that the presence of natural cracks in shale has a significant impact on stress distribution. There are two main types of acoustic emission signal distribution and evolutionary features, the evolutionary features of acoustic emission signal distribution are of two types, 0°-45° test and 60°-90° test, and the difference is mainly reflected in the damage stage, the damage of shale with high natural fracture inclination is more intense, which is manifested by the decrease in the number of acoustic emission events, but the level of acoustic emission events in the damage stage is higher, which can reach 61788, 46605 and 94315, the shale with high natural fracture inclination is more brittle.

Improvement of the Cracking Moment-Based Asphalt Mixture Splitting Test Method and Splitting Strength Research

The asphalt mixture splitting test is one of the most important methods for measuring asphalt’s tensile properties. To characterize the limitations of the traditional splitting test and the influence of the specimen size and loading conditions on the accuracy of the test, the factors affecting the strength of the splitting test were analyzed to reveal the splitting failure state and establish a unified representation model between the splitting and direct tensile tests. Initially, the moment of specimen cracking was taken as a key indicator, combined with image processing technology, to establish a set of criteria to judge the splitting test. Subsequently, standardized splitting tests were conducted and compared to tests of different specimen sizes and loading methods. Based on the octahedral strength theory, the stress points before and after the improved test were compared to the existing failure criteria. Direct tensile and splitting tests were conducted at different rates, and a unified strength–rate function model was established, realizing the unified representation of direct tensile and splitting tests. The research results indicate that the standardized splitting test is prone to the phenomenon wherein the specimen end face cracks before the center, affecting the accuracy of the test and potentially leading to redundant material strength evaluations. Using a loading method with a “35 mm specimen thickness” and a “0.3 mm rubber gasket + 12.7 mm arc-shaped batten” can essentially achieve the test hypothesis of “cracking at the center first”, resulting in less discrete outcomes and closer alignment to the three-dimensional stress failure state. The tensile and splitting strengths are both power function relationships with the rate as the independent variable, establishing a unified function model of the tensile and failure strengths. The research provides a more reliable testing method and calculation model for asphalt pavement structure design, and it also provides an effective basis for the improvement of splitting tests on materials such as concrete and rock.

A Literature Survey on Quadruped AI Assistant: Integrating Image Processing and Natural Language Processing for Emotional Intelligence

At the nexus of artificial intelligence, robotics, image processing, and natural language processing (NLP), quadruped AI assistants provide a revolutionary method of interacting with machines. Intending to give quadruped AI assistants emotional intelligence, this literature review methodically looks over and compiles the body of research, concentrating on how to use natural language processing and image processing together. The review seeks to offer a comprehensive grasp of the developments, challenges, and possible uses in this diverse subject. The initial section of the survey provides an overview of quadruped robots and highlights how they are integrated with image processing technology to provide visual perception. It explores the various locomotion techniques, demonstrating how these robots use picture data to improve their capacity to navigate and adapt to different situations. The conversation also touches on sensor technologies, highlighting their function in obtaining and deciphering visual data for intelligent interaction. In addition, the research delves into how quadruped AI assistants include natural language processing and examines how these robots interpret and react to instructions in human language. One of the main topics of debate is how sentiment analysis and emotional recognition methods might help these assistants become more emotionally intelligent. Finally, this review of the literature offers a comprehensive viewpoint on how natural language processing and image processing are integrated in quadruped AI assistants, providing academics and practitioners with an outline for developing emotional intelligence in these robots. Compiling knowledge from robotics, artificial intelligence, image processing, and natural language processing is essential to creating emotionally competent quadruped AI assistants and ensuring their smooth incorporation into human-centered settings.This research also investigates the integration of Natural Language Processing and Image Processing for Emotional Intelligence in Quadruped AI Assistants and thoroughly assesses the gyroscope functioning on these robotic systems

Using Learning Focal Point Algorithm to Classify Emotional Intelligence

Recognizing the fundamental role of learners' emotions in the educational process, this study aims to enhance educational experiences by incorporating emotional intelligence (EI) into teacher robots through artificial intelligence and image processing technologies. The primary hurdle addressed is the inadequacy of conventional methods, particularly convolutional neural networks (CNNs) with pooling layers, in imbuing robots with emotional intelligence. To surmount this challenge, the research proposes an innovative solution—introducing a novel learning focal point (LFP) layer to replace pooling layers, resulting in significant enhancements in accuracy and other vital parameters. The distinctive contribution of this research lies in the creation and application of the LFP algorithm, providing a novel approach to emotion classification for teacher robots. The results showcase the LFP algorithm's superior performance compared to traditional CNN approaches. In conclusion, the study highlights the transformative impact of the LFP algorithm on the accuracy of classification models and, consequently, on emotionally intelligent teacher robots. This research contributes valuable insights to the convergence of artificial intelligence and education, with implications for future advancements in the field.