Image Processing Approach Research Articles

Tracking of lineage mass via quantitative phase imaging and confinement in low refractive index microwells.

Measurements of cell lineages are central to a variety of fundamental biological questions, ranging from developmental to cancer biology. However, accurate lineage tracing requires nearly perfect cell tracking, which can be challenging due to cell motion during imaging. Here we demonstrate the integration of microfabrication, imaging, and image processing approaches to demonstrate a platform for cell lineage tracing. We use quantitative phase imaging (QPI), a label-free imaging approach that quantifies cell mass. This gives an additional parameter, cell mass, that can be used to improve tracking accuracy. We confine lineages within microwells fabricated to reduce cell adhesion to sidewalls made of a low refractive index polymer. This also allows the microwells themselves to serve as references for QPI, enabling measurement of cell mass even in confluent microwells. We demonstrate application of this approach to immortalized adherent and nonadherent cell lines as well as stimulated primary B cells cultured ex vivo. Overall, our approach enables lineage tracking, or measurement of lineage mass, in a platform that can be customized to varied cell types.

Multi-Angle Facial Recognition: Enhancing Biometric Security with a Broadly Positioned Stereo-Camera System

This study addresses the vulnerabilities of traditional monocular camera-based face recognition systems, emphasizing the need for improved security and reliability in biometric authentication under varying environmental conditions, lighting, and human poses. To counteract the risk of spoofing attacks using masks or static images, we introduce a multi-angle stereo camera system. This system is strategically designed to capture facial imagery from multiple perspectives, thereby enhancing depth perception and spatial accuracy, crucial for high-security authentication. Employing a novel image processing approach, the study integrates a Convolutional Neural Networks (CNN) with a simple Boolean operation to differentiate the landmarks detected on each camera. This method exploits CNN’s robust feature extraction capabilities and the effective usage of stereo camera, enabling precise detection and analysis of 3D facial landmarks. Such an approach significantly bolsters the system’s ability to differentiate between genuine faces and deceptive representations like masks or static images. Empirical results demonstrate that the stereo camera configuration substantially improves recognition accuracy, reducing both false positives and negatives, especially in controlled spoofing scenarios. The advanced 3D facial landmark detection further reinforces the system’s security. With its enhanced robustness and security, the developed system shows great potential for applications in areas requiring stringent identity verification, such as banking, public facilities, and smart home technologies.

Burst Image Restoration and Enhancement.

Burst Image Restoration aims to reconstruct a high-quality image by efficiently combining complementary inter-frame information. However, it is quite challenging since individual burst images often have inter-frame misalignments that usually lead to ghosting and zipper artifacts. To mitigate this, we develop a novel approach for burst image processing named BIPNet that focuses solely on the information exchange between burst frames and filter-out the inherent degradations while preserving and enhancing the actual scene details. Our central idea is to generate a set of pseudo-burst features that combine complementary information from all the burst frames to exchange information seamlessly. However, due to inter-frame misalignment, the information cannot be effectively combined in pseudo-burst. Thus, we initially align the incoming burst features regarding the reference frame using the proposed edge-boosting feature alignment. Lastly, we progressively upscale the pseudo-burst features in multiple stages while adaptively combining the complementary information. Unlike the existing works, that usually deploy single-stage up-sampling with a late fusion scheme, we first deploy a pseudo-burst mechanism followed by the adaptive-progressive feature up-sampling. The proposed BIPNet significantly outperforms the existing methods on burst super-resolution, low-light image enhancement, low-light image super-resolution, and denoising tasks. The pre-trained models and source code are available at https://github.com/akshaydudhane16/BIPNet.

DP Patch: ROI-Based Approach of Privacy-Preserving Image Processing With Robust Classification

DP Patch: ROI-Based Approach of Privacy-Preserving Image Processing With Robust Classification

Super-tokens Auto-encoders for image compression and reconstruction in IoT applications

New telecommunications networks are enabling powerful AI applications for smart cities and transport. These applications require real-time processing of large amounts of media data. Sending data to the cloud for processing is very difficult due to latency and energy constraints. Lossy compression can help, but traditional codecs may not provide enough quality or be efficient enough for resource-constrained devices. This paper proposes a new image compression and processing approach based on variational auto-encoders (VAEs). This VAE-based method aims to efficiently compress images while still allowing for high-quality reconstruction and object detection tasks. The encoder is designed to be lightweight and suitable for devices with limited computing power. The decoder is more complex and uses multi-level vector quantization to reconstruct high-resolution images. This approach allows for a simple encoder on edge devices and a powerful decoder on cloud servers. Key contributions include a low-complexity encoder, a new VAE model based on vector quantization, and a framework for using VAEs in IoT. The first experiments on reconstructed images on CelebA and ImageNet100 datasets show promising results in terms of MS-SSIM, PSNR, MSE and rFID compared to the literature and the ability of our approach to be used in IoT applications. Our approach presents results similar to complex algorithms like compression algorithms BPG in term of trade-off rate-distortion, and hierarchical auto-encoder (HQA) in terms of image reconstruction quality

Revolutionizing Two-Dimensional Pattern Arrangement: Innovative Image Processing Approaches

Revolutionizing Two-Dimensional Pattern Arrangement: Innovative Image Processing Approaches

Machine learning-based morphological and mechanical prediction of kirigami-inspired active composites

Kirigami-inspired designs hold great potential for the development of functional materials and devices, but predicting the morphological configuration of these structures under various loading conditions remains a challenge for traditional experimental and numerical methods. Here, we present a novel approach that utilizes machine learning algorithms to accurately predict the deformation and stress field of kirigami-inspired programmable active composites. To train our model, first, we used a chemical corrosion algorithm to generate a dataset of kirigami-inspired imaging model accompanied by utilizing finite element simulations to obtain their deformation and stress fields as the ground truth, and subsequently trained the machine learning model to offer robust predictions of the displacement and stress fields of the designated structures. The graphically preprocessing transformation between color space and deformation(stress) fields is used to match the fields prediction of mechanical problems with powerful machine learning approaches in image processing. Our results demonstrate the effectiveness of this approach in predicting the morphological and mechanical behaviors of kirigami-inspired active structures, paving the way for the development of advanced and functional composite designs that are programmable and active.

Maturity stages detection prototype device for classifying custard apple (Annona squamosa L) fruit using image processing approach

Custard apple (Annona squamosa L.) is commercially grown for its sweetness, delicacy, and nutraceutical benefits across the world. Fruit must be harvested from trees before the climacteric stage begins for stabilizing the maturation process and perishability, which otherwise restricts longer shelf–life. Therefore, maturity–based grading or classifying custard apple fruit in accordance with their ripening is required for commercial marketing and industrial storage purposes. This is a crucial task that agriculturists and the food processing industries must have to perform manually, which is ineffective and susceptible to human errors. In this research, we have proposed an image processing–based maturity stages detection prototype device for fruit classification of custard apple (Cv. Balanagar) based on changes in their physical, chemical and imaging features (skin color of fruit areoles). The newly developed handheld device comprises of a Raspberry Pi board, camera module, LCD touch screen, and image processing algorithms to identify five essential maturity stages i.e. 0 %, 25 %, 50 %, 75 % and 100 % areoles opening of custard apple. The images were pre–processed and segmented, after which imaging aspects of custard apple fruit including relative R, G and B components were extracted and their relationship with chemical properties such as TSS was established. For classification of captured images, K–means clustering and SVM algorithms were performed on Raspberry Pi Platform with model training and classifier codes in Python (3.9.6) to get desired results. The statistical analysis revealed a substantial difference between the maturity stages in terms of R, G and TSS. Overall, 100 % accuracy was accomplished with respect to the grading of custard apple fruit at varied maturity. This could be a low–cost device useful for farm–level custard apple fruit classification, minimising economic losses caused by faulty supply entering the marketing chain. With slight modifications in algorithms, this device could also be used to assess the maturity of other fruit crops. It could be also an alternative to several expensive approaches being used for automating the maturity grading process that is not feasible, particularly for farmers working in field conditions.

An Improved Inception Network to classify black tea appearance quality

To conveniently and precisely evaluate orthodox black tea appearance quality, we here present a novel method based on computer vision integrated image processing and deep learning. Tea images were collected using the custom-built image acquisition device and processed with Adaptive Local Tone Mapping (ALTM), and Unsharp Masking to optimize illumination and sharpness. These images were then used to train six convolutional neural networks (CNNs) to find suitable network structure by transfer learning. It was found that the CNNs constructed with the MBconv modules had better performance in this task. Consequently, a CNN classification model (Improved Inception Network) based on the MBconv modules was constructed and trained, which yielded a test accuracy of 95%, performed better than the other CNNs; the test accuracy of original Inception V3 was 89%. The proposed method obtained 97.22% accuracy for independent set in validation, which demonstrated the viability of applying image processing and deep learning approaches to solve practical problems in the field of tea assessment.

STEAM: Spatial Trajectory Enhanced Attention Mechanism for Abnormal UAV Trajectory Detection

Accurate unmanned aerial vehicle (UAV) trajectory tracking is crucial for the successful execution of UAV missions. Traditional global positioning system (GPS) methods face limitations in complex environments, and visual observation becomes challenging with distance and in low-light conditions. To address this challenge, we propose a comprehensive framework for UAV trajectory verification, integrating a range-based ultra-wideband (UWB) positioning system and advanced image processing technologies. Our key contribution is the development of the Spatial Trajectory Enhanced Attention Mechanism (STEAM), a novel attention module specifically designed for analyzing and classifying UAV trajectory patterns. This system enables real-time UAV trajectory tracking and classification, facilitating swift and accurate assessment of adherence to predefined optimal trajectories. Another major contribution of our work is the integration of a UWB system for precise UAV location tracking, complemented by our advanced image processing approach that includes a deep neural network (DNN) for interpolating missing data from images, thereby significantly enhancing the model’s ability to detect abnormal maneuvers. Our experimental results demonstrate the effectiveness of the proposed framework in UAV trajectory tracking, showcasing its robust performance irrespective of raw data quality. Furthermore, we validate the framework’s performance using a lightweight learning model, emphasizing both its computational efficiency and exceptional classification accuracy.

RepDNet: A re-parameterization despeckling network for autonomous underwater side-scan sonar imaging with prior-knowledge customized convolution

Side-scan sonar (SSS) is now a prevalent instrument for large-scale seafloor topography measurements, deployable on an autonomous underwater vehicle (AUV) to execute fully automated underwater acoustic scanning imaging along a predetermined trajectory. However, SSS images often suffer from speckle noise caused by mutual interference between echoes, and limited AUV computational resources further hinder noise suppression. Existing approaches for SSS image processing and speckle noise reduction rely heavily on complex network structures and fail to combine the benefits of deep learning and domain knowledge. To address the problem, RepDNet, a novel and effective despeckling convolutional neural network is proposed. RepDNet introduces two re-parameterized blocks: the Pixel Smoothing Block (PSB) and Edge Enhancement Block (EEB), preserving edge information while attenuating speckle noise. During training, PSB and EEB manifest as double-layered multi-branch structures, integrating first-order and second-order derivatives and smoothing functions. During inference, the branches are re-parameterized into a 3 × 3 convolution, enabling efficient inference without sacrificing accuracy. RepDNet comprises three computational operations: 3 × 3 convolution, element-wise summation and Rectified Linear Unit activation. Evaluations on benchmark datasets, a real SSS dataset and Data collected at Lake Mulan aestablish RepDNet as a well-balanced network, meeting the AUV computational constraints in terms of performance and latency.

Early detection of diabetes potential using cataract image processing approach

Diabetes is a disease characterized by a high level of sugar in the blood. The disease occurs because of a disruption in the metabolic system when insulin is not produced effectively and functions properly. High blood sugar levels, for an extended period of time, can harm a few organ systems, including the heart and kidneys. Moreover, it may cause blindness or death if it is not carefully monitored. Because diabetes symptoms are rarely seen, one of the factors that may cause diabetes is self-awareness. Thus, with Artificial Intelligence, this problem can be solved. Artificial intelligence studies how machines can function like humans. This study implemented a Convolutional Neural Network algorithm with (1) input layer, (2) feature learning layer, (3) classification layer, and (4) output layer as the architecture for AI. The accuracy of the developed AI model was measured from its precision, recall, and f1-score. The results show that the model obtained 90% precision, recall, and f1-score for real-world cases found in two hospitals located in Solo and Yogyakarta, Indonesia. According to the results of the tests, 9 out of 10 patients were correctly predicted as having a high risk of diabetes based on their eye images.

AAPM Task Group Report 311: Guidance for performance evaluation of fluorescence-guided surgery systems.

The last decade has seen a large growth in fluorescence-guided surgery (FGS) imaging and interventions. With the increasing number of clinical specialties implementing FGS, the range of systems with radically different physical designs, image processing approaches, and performance requirements is expanding. This variety of systems makes it nearly impossible to specify uniform performance goals, yet at the same time, utilization of different devices in new clinical procedures and trials indicates some need for common knowledge bases and a quality assessment paradigm to ensure that effective translation and use occurs. It is feasible to identify key fundamental image quality characteristics and corresponding objective test methods that should be determined such that there are consistent conventions across a variety of FGS devices. This report outlines test methods, tissue simulating phantoms and suggested guidelines, as well as personnel needs and professional knowledge bases that can be established. This report frames the issues with guidance and feedback from related societies and agencies having vested interest in the outcome, coming from an independent scientific group formed from academics and international federal agencies for the establishment of these professional guidelines.

Evolutionary meta-heuristic optimized model: An application to plant disease diagnosis

Deep learning strategies have been achieved over the historical decades to resolve many computer vision applications. Recently, these deep learning algorithms have been extensively used as a tool in classification problems. Generally, the deep learning algorithms trained with gradient-based optimizers, which has some downsides such as the slow speed of convergence and stuck in local minima. As a solution, the planned work using meta-heuristic based Grey Wolf and Whales optimization algorithms for the automatic plant disease detection model. The planned work has explored the application of automatic plant disease identification through the leaf images with the help of the image processing approach. The planned research has evaluated the deep learning algorithm with Grey Wolf and Whales optimization techniques using the three types of datasets, such as Plant Village, New Plant Disease, and Rice Leaf Disease databases. The simulation consequences illustrate that the computational efficiency of the Grey Wolf and Whales based automatic disease identification process is boosted when coupled with the deep learning method.

An image processing approach to reconstruct wind using long-range wind lidars

Lidar-based wind sensing technology, originally used in the wind energy sector, is now being utilized in wind engineering to monitor wind action for designing fjord-crossing infrastructure like long-span bridges. Accurate estimation of design wind loads is crucial for the design process of such bridges. This paper examines wind data from two pairs of long-range lidars positioned along one side of the Sulafjorden, Norway, in a measurement campaign initiated by the Norwegian Public Roads Administration (NPRA). Two different scanning modes, the Plan Position Indicator (PPI) mode and the staring mode, with the fixed line-of-sight (LOS) orientation, are used to determine the line-of-sight wind speeds, the wind direction, and the related along wind speed. An image processing approach is used to compute along wind velocity information from all range gates based on the LOS wind speeds from two near-parallel lidar beams. The estimated along wind velocity is validated through wind data calculated at the intersection point of the two lidars, which provides velocity data with two horizontal components. The image-based reconstruction is found to produce reasonable wind speed estimates with a goodness-of-fit coefficient of R2 = 0.903. The mean wind direction estimate is smoothened but still comparable to the actual wind direction. The image processing approach shows promising potential to provide wind speed and direction information for all range gates for the lidar setup used, which can supplement traditional single-point wind velocity characterization by dual lidars.

Automated Plant Disease Diagnosis Using Deep Learning Model

With the world's population growing, it is vital that food crops and medicinal plants are required to be developed in large quantity. A model that can help us understand the process more comprehensively will be explored. Based on a thorough understanding, we are exploring a model that will enable us to extract the precise information and in-depth knowledge we require. There are many plants that contract lethal diseases every year, including food plants. This reduces their output rates. We must employ automated methods to locate the issue inside the facility if we do not want manufacturing prices to rise. The detection of plant diseases can be automated through robotics and image processing. As a result of recent technological advancements, we now have the ability to simplify our artwork. A deep learning and image processing approach can increase the efficiency of detection processes. A great deal of progress has been made in diagnosing plant diseases. An artificial intelligence model was trained to assist in diagnosing flora disorders. Based on the photo information (leaf), the model can be controlled. Identifying the problem and discovering the condition in this study, we propose the use of computer vision technology combined with fuzzy logic to detect and grade leaf diseases. GLCM is performed to extract texture features, and fuzzy logic is applied to grade the disease. K-means clustering is applied for determining defected areas; GLCM is used to determine defected areas; and fuzzy logic is used for diagnosing diseases. About 70% of classifications are accurate according to the model. By using Speed Up Robust Features (SURF), DENSE, and Bag of Visual Words (BOVW) in addition to the global features, the accuracy of the system can be enhanced. A treatment program will help people better understand the illness if one is offered.

Morphometric Analysis of Chili Leaves with Yellow Curly Symptom Using Digital Image Processing Approach and Data Mining Algorithm

Gejala kuning keriting pada daun cabai umumnya disebabkan oleh infeksi Begomovirus. Daun tanaman terinfeksi tidak hanya mengalami perubahan warna sebagai indikator rusaknya klorofil tetapi juga mengalami perubahan morfologi bentuk. Penelitian ini bertujuan menguantifikasi gejala infeksi Begomovirus berdasarkan perubahan morfologi bentuk daun menggunakan pengolahan citra digital dan algoritma data mining yang akan memudahkan dalam pemantauan dan analisis perkembangan penyakit tanaman. Total 33 citra daun cabai rawit bergejala kuning keriting maupun tidak bergejala menjadi dataset penelitian ini. Citra daun cabai tersebut diolah dan diekstrak karakteristik bentuknya berupa circularity, aspect ratio, roundness, dan solidity menggunakan aplikasi Fiji-ImageJ. Selanjutnya dilakukan uji beda (uji-t), pengelompokan citra menggunakan algoritma Simple K-Means, dan evaluasi ketepatan hasil pengelompokan berdasarkan indeks ARI dan NMI. Hasil penelitian menunjukkan bahwa secara umum ada perbedaan bentuk yang nyata antara daun bergejala dengan daun tidak bergejala. Daun cabai rawit bergejala kuning keriting memiliki rata-rata nilai aspect ratio dan solidity yang lebih kecil dibandingkan daun cabai tidak bergejala, sebaliknya memiliki rata-rata nilai circularity dan roundness yang lebih besar dibandingkan daun cabai tidak bergejala. Evaluasi ketepatan pengelompokan sampel daun cabai rawit bergejala maupun tidak bergejala berdasarkan indeks ARI dan NMI menghasilkan nilai terbaik untuk pengelompokkan ke dalam dua kelompok.

Novel Feature Extraction for Oil Palm Bunches Classification

This paper presents research on an image processing approach for oil palm fruit brunches classification on automating the process of classifying palm fruit brunches based on their visual characteristics, which can have applications in the agriculture and palm oil industry. The research aims to classify the bunches into four categories which are unripe, under ripe, ripe and over ripe. Additionally, this approach can reduce the manual labour involved in palm oil grading and provide a more objective and consistent method for grading the oil. The proposed approach consists of the process of image acquisition, image pre-processing, colour processing, image segmentation and classification. The ripeness of the oil palm fruit bunches is determined based on the percentage of ripeness areas masked on the fruit surface. The proposed algorithms successfully classified the palm oil bunches and improved the accuracy of grading palm oil, achieving 85% accuracy from the experiment results.

Automated models of visual information processing

The article presents a study devoted to the development and research of an automated model of visual information processing. The goal of the research was to create a comprehen-sive model capable of automatically processing and analyzing various forms of visual data, such as images and videos. The model is developed on the basis of a combined approach that combines various algorithms and methods of visual information processing. The literature review conducted within the scope of this study allowed us to study the existing methods and algorithms for visual information processing. Various image processing approaches were analyzed, including segmentation, pattern recognition, object classification and detection, video analysis, and other aspects. As a result of the review, the advantages and limitations of each approach were identified, as well as the areas of their application were determined. The developed model showed high accuracy and efficiency in visual data processing. It can suc-cessfully cope with the tasks of segmentation, recognition and classification of objects, as well as video analysis. The results of the study confirmed the superiority of the proposed model. Potential applications of the automated model are considered, such as medicine, robotics, security, and many others. However, limitations of the model such as computational resource requirements and quality of input data are also noted. Further development of this research can be aimed at optimizing the model, adapting it to specific tasks and expanding its func-tionality. In general, the study confirms the importance of automated models of visual infor-mation processing and its important place in modern technologies. The results of the research can be useful for the development of new systems based on visual data processing and con-tribute to progress in the field of computer vision and artificial intelligence.

Multi-omics fusion with soft labeling for enhanced prediction of distant metastasis in nasopharyngeal carcinoma patients after radiotherapy

Omics fusion has emerged as a crucial preprocessing approach in medical image processing, significantly assisting several studies. One of the challenges encountered in integrating omics data is the unpredictability arising from disparities in data sources and medical imaging equipment. Due to these differences, the distribution of omics futures exhibits spatial heterogeneity, diminishing their capacity to enhance subsequent tasks. To overcome this challenge and facilitate the integration of their joint application to specific medical objectives, this study aims to develop a fusion methodology for nasopharyngeal carcinoma (NPC) distant metastasis prediction to mitigate the disparities inherent in omics data. The multi-kernel late-fusion method can reduce the impact of these differences by mapping the features using the most suiTable single-kernel function and then combining them in a high-dimensional space that can effectively represent the data. The proposed approach in this study employs a distinctive framework incorporating a label-softening technique alongside a multi-kernel-based Radial basis function (RBF) neural network to address these limitations. An efficient representation of the data may be achieved by utilizing the multi-kernel to map the inherent features and then merging them in a space with many dimensions. However, the inflexibility of label fitting poses a constraint on using multi-kernel late-fusion methods in complex NPC datasets, hence affecting the efficacy of general classifiers in dealing with high-dimensional characteristics. The label softening increases the disparity between the two cohorts, providing a more flexible structure for allocating labels. The proposed model is evaluated on multi-omics datasets, and the results demonstrate its strength and effectiveness in predicting distant metastasis of NPC patients.