Image Techniques Research Articles

Crow search algorithm with deep transfer learning driven skin lesion detection on dermoscopic images

The early diagnosis of skin cancer is of paramount importance for effective patient treatment. Dermoscopy, a non-surgical technique, utilizes precise equipment to examine the skin and plays a crucial role in identifying specific features and patterns that may indicate the presence of skin carcinoma. In recent times, machine learning (ML) methods have been developed to recognize and classify dermoscopic images as either malignant or benign. Deep learning (DL) systems, including Convolutional Neural Networks (CNNs), as well as various ML models like Random Forest (RF) classifiers and Support Vector Machine (SVM), are employed to extract relevant features from these images. This study introduces the Crow Search Algorithm with Deep Transfer Learning Driven Skin Lesion Detection on Dermoscopic Images (CSADTL-SLD) technique. The CSADTL-SLD method starts with the application of a median filter (MF) to remove noise from the images and utilizes the GoogleNet model for feature extraction. GoogleNet is well-regarded for its capacity to capture intricate and meaningful patterns within the data, which are essential for accurate lesion characterization. Furthermore, the CSADTL-SLD technique applies the Crow Search Algorithm (CSA) for parameter tuning of the GoogleNet model. After feature selection, the system employs the MLP classification model for precise lesion categorization. The comprehensive results of this research demonstrate the superiority of the CSADTL-SLD algorithm, showing significant enhancements in skin lesion detection accuracy and robustness when compared to existing methods. This approach holds promise as an effective solution for automating the detection and classification of skin lesions in dermoscopic images.

MRI-based synthetic CT for assessment of the bony elements of the sacroiliac joints in children

ObjectivesThe purpose of this study is to assess the equivalency of MRI-based synthetic CT (sCT) to conventional CT for sacroiliac joint bony morphology assessment in children.MethodsA prospective study was performed. Children who had (PET-)CT-scan underwent additional MRI. sCT-CT image quality was analyzed by two readers subjectively overall, semi-quantitatively in terms of cortical delineation, joint facet defects, growth plate fusion, ossified nuclei, lumbosacral transitional anomaly, and bony bridges, and quantitatively for disc space height, spinal canal width, and sacral vertebrae width and height. Cohen’s kappa and equivalence analyses with Bland–Altman plots were calculated for categorical and continuous measures respectively.ResultsTen patients were included (6 boys; aged 9–16 years; mean age 14 years). Overall sCT image quality was rated good. Semi-quantitative assessment of cortical delineation of sacroiliac joints, bony bridges, and joint facet defects on the right iliac and sacral sides showed perfect agreement. Correlation was good to excellent (kappa 0.615–1) for the presence of lumbosacral transitional anomaly, fusion of sacral growth plates, joint facet defect, and presence of ossified nuclei. sCT-CT measurements were statistically equivalent and within the equivalence margins (–1–1 mm) for intervertebral disc space height and spinal canal width.Intra- and inter-reader reliability was excellent for quantitative assessment (0.806 < ICC < 0.998). For categorical scoring, kappa ranged from substantial to excellent (0.615–1).ConclusionsCT appears to be visually equivalent to CT for the assessment of pediatric sacroiliac joints. sCT may aid in visualizing sacroiliac joints compared to conventional MRI, with the benefit that no ionizing radiation is used, especially important in children.Critical relevance statementMRI-based synthetic CT, a new technique that generates CT-like images without ionizing radiation, appears to be visually equivalent to CT for assessment of normal pediatric sacroiliac joints and can potentially assess structural damage as it clearly depicts bony cortex.Key points• MRI-based sCT is a new image technique that can generate CT-like images.• We found that sCT performs similarly to CT in displaying bony structures of pediatric sacroiliac joints.• sCT has already been clinically validated in the sacroiliac joints in adults.• sCT can potentially assess structural damage from erosions or ankylosis as it clearly depicts bony cortex.Graphical

Investigation of patients with new infection of echinococcal cyst in Sardinia, Italy

Cystic Echinococcosis (CE) is a zoonotic disease caused by the larval stage of a tapeworm of Taeniidae family, genus Echinococcus and species Echinococcus granulosus sensu lato (s.l.). CE is a worldwide public health problem and is highly incident in all Mediterranean areas.Following clinical, image techniques and serological investigations all 83 subjects involved in the study were diagnosed for CE. General and clinical data were entered into a database and evaluated. The 43.37% were female and 56.63% male, mean age was 50.71 while the range most represented (22.7%) was between 61->70 years. The purposes of our survey were to investigate these 83 patients enrolled in the study and to deeply examine 20 (24.10%) patients that developed a new echinococcal cyst. Moreover, the causes at the basis of the onset of a new cyst were analysed, together with a possible correlation with different treatments related to primary infection corresponding to surgery (n=7), albendazole (n=6), PAIR (n=3) and watch and wait (n=4). A possible link with medical treatments of the primary infection was observed in the subjects who underwent surgery or PAIR and a likely correlation attributable to high environmental contamination in subjects managed with drugs or watch and wait was detected. Moreover, our analysis evidenced that patients with a new infection presented a more severe diagnosis along with a major mortality rate. Finally, these data may have an important contribution for an epidemiological point of view concerning the percentage of CE in a specific geographical endemic area, such as Sardinia.

Stacked neural nets for increased accuracy on classification on lung cancer

Lung cancer is regarded as one of the most lethal diseases endangering human survival. It is difficult to detect lung cancer in its early stages, because of the ambiguity in the lung regions in the medical images. Healthcare business is automating itself with the use of image recognition and data analytics, much as the computing sector has completely automated. This article proposes a novel architecture, the Stacked Neural Network (SNN), for the detection and classification of lung cancer using CT scan data. The goal of the proposed technique is to investigate the accuracy levels of different Neural Networks (NN) and determine the early stage of lung cancer. The most effective technique for processing medical images, classifying and detecting lung nodules, extracting features, and predicting the stage of lung cancer is deep learning. First, lung areas are extracted using image processing techniques. SNN is used for the segmentation process. Various neural network techniques are utilised for the classification process once the features are retrieved from the segmented pictures. The suggested methods' performances are assessed using F1-Measure, accuracy, precision, and recall metrics. 96% classification accuracy is shown in the testing findings, which is comparatively greater than other methods currently in use. Proposed algorithm is clearly supported for real-world clinical practice.

Natural surface floaters in image-based river surface velocimetry: Insights from a case study

This study focuses on utilizing image techniques for river velocity measurement, with a specific emphasis on natural surface floating patterns. Employing a multi-camera system, we conducted 3D measurements on river surfaces, including surface velocity and water surface reconstruction. A pattern-based tracking approach has been adopted to improve the performance of image measurements on different types of natural floating tracers. The study employs the following approaches: 3D Lagrangian Pattern Tracking Velocimetry (3D-LPTV), 2D Lagrangian Pattern Velocimetry (2D- LPTV), and Large-scale Particle Image Velocimetry (LSPIV), for surface velocity estimation. The outcomes revealed that all three approaches yielded consistent results in terms of averaged velocity. However, the LSPIV method produced about two times higher uncertainty in measured velocities compared to the other methods. A strategy to assess the quality of river surface patterns in velocity estimation is presented. Specifically, the sum of squared interrogation area intensity gradient (SSIAIG) was found to be strongly correlated with measurement uncertainty. Additionally, a term related to the peak sidelobe ratio (PSR) of the cross-correlation map was found as an effective constraint, ensuring the image-tracking process achieves high reliability. The precision of measurements increases corresponding to the increase of image intensity gradient and PSR.

A Light Weight Depthwise Separable Layer Optimized CNN Architecture for Object-Based Forgery Detection in Surveillance Videos

Abstract The present era is at the peak of technological advancement in image and video processing techniques, with user-friendly accessible tools/techniques. This immersive technology development makes video forensics enormously challenging. Specifically, the passive approaches to object-based forgeries in videos are crucial for legal and judicial matters. Hence, to ensure the integrity of the videos, a scientific, statistical and passive investigation of videos is required to maintain the spatial and temporal information content. This paper aims to develop a passive approach for digging out the forgery traces by applying the motion residue windowing technique for object removal forgery in surveillance videos. The novel max averaging windowing techniques improve visual imprints of the object removal forgery in the videos from the existing methods in the literature. A deep learning approach is the next step for achieving forgery detection in surveillance videos. The proposed lightweight depth-separable layer-optimized CNN has fast execution speed, optimized in terms of parameters without compromising the desired accuracy. This network is trained at a frame level with 98.60% testing accuracy, followed by a pipeline architecture of the proposed model for detection of forgery at video level with 99.01% accuracy. The suggested model works better than current models regarding post-processing operations, compression rates, forged video detection accuracy, precision, recall and F1 score.

War, image, art: From vision to judgement

There are excellent research papers in the field of visual studies that examine the relationship between war and images. This paper has other and additional aims. The first is to examine not so much how war is transferred from the ground to image production, but how war, as intrusion of the real, forces a general reflection on image techniques. The second is to examine whether there is an instance of art that is somehow different from the instance of the mere image, which is always related to something that manifests itself in war, and which, especially on the occasion of a war, invites a broader reflection on existence, judgement and the world. The paths taken for these investigations are rooted in the idea that a (political) theory of art cannot be reduced to an internal classification of image theory, since it investigates aspects that go beyond the definition of the image itself.

Quantum ghost imaging of a vector field.

Quantum ghost image technique utilizing position or momentum correlations between entangled photons can realize nonlocal reconstruction of the image of an object. In this work, based on polarization entanglement, we experimentally demonstrate quantum ghost imaging of vector images by using a geometric phase object. We also provide a corresponding theoretical analysis. Additionally, we offer a geometrical optics path explanation of ghost imaging for vector fields. The proposed strategy offers new insights into the fundamental development of ghost imaging and also holds great promise for developing complex structured ghost imaging techniques. Our work expanding the principle of ghost imaging to spatially varying vector beams will lead to interesting developments of this field.

Speckle Reduction in Digital Holography by Fast Logistic Adaptive Non-Local Means Filtering

Digital holography is a promising imaging technology. However, there is speckle noise in the reconstructed image of a digital hologram. Speckle degrades the quality of the reconstructed image. Suppression of speckle noise is a challenging problem in digital holography. A novel method is proposed to reduce speckle by a fast logistic adaptive non-local means (LA-NLM) algorithm. In the proposed method, the logistic function is incorporated into the weight calculation of the NLM algorithm to account for multiplicative speckle noise. Filtering parameters are dynamically adjusted according to the statistical property of speckle in the reconstructed image. To enhance computational efficiency, the proposed algorithm takes advantage of the integral image technique to speed up the calculation of the similarity between image patches. Simulated and experimental digital holograms are obtained to verify the proposed method. The results show that the speckle noise is effectively suppressed in digital holography. The proposed method is efficient and feasible, and can be applied to such fields as three-dimensional display, holographic measurement, and medical diagnosis.

Thorax X-Ray Image Segmentation Technique Using Four Thresholding Algorithm Variants

Pneumonia is a respiratory infection caused by bacteria, viruses or fungi, and has been recognized as a fairly common and threatening disease. When diagnosing this disease, doctors usually also use thorax X-ray images. Nowadays, diagnosing pneumonia has been made possible with the help of machine learning technology. Doctors or medical personnel in locations where there are no pulmonary specialists or experts can be assisted by this technology. Machine learning algorithms are used to process digital images that have passed the pre-processing and segmentation stages. This paper offers a solution to segmentation technique of thorax X-ray digital image using a combination of four thresholding algorithms. This combination aims to find the best CNN model with segmentation techniques in the form of the most suitable thresholding algorithm. The result of this research is four different data sets. The thresholding algorithms used include binary, thresh binary inv, thresh to zero, thresh tozero inv with a threshold value of 150. The data used in this research is a thorax X-ray image dataset, as many as 5,856 images acquired from the Kaggle repository data. The program code in this research uses the Python programming language in the Anaconda environment. This research has resulted in a comparison of the accuracy values obtained using 4 variants between thres_binary thresholding algorithm and thres_binary_inv. The thres_tozero obtained 95% of accuracy while thres_tozero_inv obtained 94% of accuracy.

Failure Characterization of Al-Zn-Mg Alloy and Its Weld Using Integrated Acoustic Emission and Digital Image Techniques

The three-point bending damage process of an A7N01 aluminum alloy body material and weld seam in an electric multiple unit train was monitored using acoustic emission (AE) and digital image technology. The AE signal characteristics of static load damage to the aluminum alloy and weld seam were studied using the AE signal parameter and time–frequency analysis. Based on the observation of the microstructure and fracture morphology, the source mechanism of AE signals was analyzed. The experimental results indicate that AE energy, centroid frequency, and peak frequency are effective indices for predicting the initiation of cracks in A7N01 aluminum alloy and weld seams. The digital image monitoring results of the notch tip damage evolution of aluminum alloy samples confirmed the predictions based on AE energy, centroid frequency, and peak frequency for crack initiation. The AE signal source mechanism revealed that the differences in AE characteristics between the base material and weld seam can be attributed to microstructure variations and fracture modes. In summary, the AE technique is more sensitive to changes in the fracture mode and can be utilized to monitor the damage evolution of welded structures.

An introduction to double stain normalization technique for brain tumour histopathological images

Stain normalization is an image pre-processing method extensively used to standardize multiple variances of staining intensity in histopathology image analysis. Staining variations may occur for several reasons, such as unstandardized protocols while preparing the specimens, using dyes from different manufacturers, and varying parameters set while capturing the digital images. In this study, a double stain normalization technique based on immunohistochemical staining is developed to improve the performance of the conventional Reinhard’s algorithm. The proposed approach began with preparing a target image by applying the contrast-limited adaptive histogram equalization (CLAHE) technique to the targeted cells. Later, the colour distribution of the input image will be matched to the colour distribution of the target image through the linear transformation process. In this study, the power-law transformation was applied to address the over-enhancement and contrast degradation issues in the conventional method. Five quality metrics comprised of entropy, tenengrad criterion (TEN), mean square error (MSE), structural similarity index (SSIM) and correlation coefficient were used to measure the performance of the proposed system. The experimental results demonstrate that the proposed method outperformed all conventional techniques. The proposed method achieved the highest average values of 5.59, 3854.11 and 94.65 for entropy, TEN, and MSE analyses.

Efficient color image steganography based on new adapted chaotic dynamical system with discrete orthogonal moment transforms

In today's rapidly evolving digital landscape, safeguarding the secure transmission of sensitive information is paramount. Traditional steganography methods, while effective at data concealment, have grown vulnerable to brute force attacks in our age of powerful computing. This paper addresses the pressing need to bolster data transmission security. To address the challenges posed by limited key space and security concerns associated with 1-D chaotic maps in data transmission, we introduce a novel 1-D chaotic map, the Multi-Parameter Chebyshev Map (MCM). The MCM boasts four control parameters, each with a broad chaotic range and high chaos level, significantly enhancing security. Furthermore, we present an innovative steganography technique for color images that leverages the MCM in conjunction with discrete moment transforms. This method chaotically embeds the secret message within the moment coefficients of a cover color image using MCM and quantization rules. The expansive chaotic range and high chaos level of the MCM substantially reinforce the security of this approach. The considered moment transforms include Tchebichef, Krawtchouk, Hahn, Charlier, Meixner, Shmaliy, Racah, and Dual-Hahn. The performance of the proposed steganography method is thoroughly evaluated and compared with other state-of-the-art methods in terms of hiding capacity and image quality in both spatial and transform domains. The quantization of moment coefficients demonstrates its superior performance. This method holds crucial implications for practical applications requiring secure communication of color images, such as military and medical contexts. The combination of MCM and moment transforms in this method presents a promising avenue for future research in the field of information hiding.

Economic feasibility of Gabal Um Takha leucogranitic intrusion, South Sinai, Egypt: Integrated remote sensing, geochemical, aeromagnetic, and geotechnical approach

Variable granitic rocks are widely used as decorative stones. The present work conducted accurate mapping and the utilization of a large leucogranitic intrusion (Um Takha area) as a decorative stones through physico-mechanical tests, mineralogical and geochemical compositions. These leucogranites are medium-grained, classified as syenogranites, peraluminous, calc-alkaline, and highly evolved granites developed during the post-collisional setting. This study utilized various remote sensing techniques to identify the Um Takha leucogranitic using ASTER and Landsat-9 data. A novel FCC band ratio of the ASTER image (RGB: b1+b3/b2, b2/b1, b7+b9/b8) was created, along with the Landsat-9 FCC band ratio of (4/7, 3/4, and 2/1) in the RGB, to effectively map the distribution of in the study area. These results, combined with field observation, led to the creation of a new modified geological map of the Um Takha area. The analysis of magnetic data reveals that the depth of magnetic sources under the study area ranges between 170 m and 2215 m, as calculated by source parameter image and Euler deconvolution techniques. The application of horizontal gradient, tilt derivatives and improved logistic techniques reveals that the major structural trend affecting the study area is in the N–S, NNE-SSW and NNW-SSE direction. The physico-mechanical properties of leucogranitic rocks (M13) fulfilled the requirements of dimension stone in terms of bulk density, compressive strength, and to some extent water absorption. Therefore, they could be acceptable for interior use and the light-duty purposes of exterior use, such as facades and cladding. This suggest that a large resources of leucogranitic can be used as a decorative stones.

Ensemble Learning-Based Solutions: An Approach for Evaluating Multiple Features in the Context of H&amp;E Histological Images

In this paper, we propose an approach based on ensemble learning to classify histology tissues stained with hematoxylin and eosin. The proposal was applied to representative images of colorectal cancer, oral epithelial dysplasia, non-Hodgkin’s lymphoma, and liver tissues (the classification of gender and age from liver tissue samples). The ensemble learning considered multiple combinations of techniques that are commonly used to develop computer-aided diagnosis methods in medical imaging. The feature extraction was defined with different descriptors, exploring the deep learning and handcrafted methods. The deep-learned features were obtained using five different convolutional neural network architectures. The handcrafted features were representatives of fractal techniques (multidimensional and multiscale approaches), Haralick descriptors, and local binary patterns. A two-stage feature selection process (ranking with metaheuristics) was defined to obtain the main combinations of descriptors and, consequently, techniques. Each combination was tested through a rigorous ensemble process, exploring heterogeneous classifiers, such as Random Forest, Support Vector Machine, K-Nearest Neighbors, Logistic Regression, and Naive Bayes. The ensemble learning presented here provided accuracy rates from 90.72% to 100.00% and offered relevant information about the combinations of techniques in multiple histological images and the main features present in the top-performing solutions, using smaller sets of descriptors (limited to a maximum of 53), which involved each ensemble process and solutions that have not yet been explored. The developed methodology, i.e., making the knowledge of each ensemble learning comprehensible to specialists, complements the main contributions of this study to supporting the development of computer-aided diagnosis systems for histological images.

The manifesto in Martin Scorsese's auteur films

Relevance. In the period from 1917 to the early 1960s, there was a paradigm of �classic Hollywood�, in which films were produced according to the dominant aesthetic, genre and narrative formulas, the characters represented themselves as specific typical images with understandable motivations for the general public. Martin Scorsese is a representative of cinematography, who changed classical views of film art. That is why the study of Martin Scorsese's work remains relevant for researchers in the field of cinematography and culture to the present day.Purpose. The purpose of this work is to study the creativity of director, screenwriter and actor Martin Scorsese, as well as to identify the author's style in the artist's work, determine his author's handwriting and manner.Methodology. The methodology of this research is based on theoretical methods of scientific knowledge, in particular, the method of information analysis and synthesis, the cultural method, as well as historical and comparative methods were used.Results. The study was conducted theoretical analysis of scientific literature. The work considered the life and creative path of the director. In particular, the still image technique used by Scorsese was analysed, as well as his own director's handwriting. Scorsese's activities during public speeches and discussions related to the preservation of the National Film Foundation and the development of a new cinema are considered. Such films of the director as �Taxi Driver�, �Raging Bull�, �Nice Guys�, �Cape Fear�, �Age of Innocence�, �Casino�, �Gangs of New York�, �Aviator�, �The Departed�, �Island of the Damned�, �Silence� were analysed.Conclusions. The results of the research can be used by students of directing specialties and teachers of special disciplines to study the creativity of the director, screenwriter and actor Martin Scorsese during theoretical practical classes.

Retracted: Multisensor Data and Cross-Validation Technique for Merging Temporal Images for the Agricultural Performance Monitoring System

Retracted: Multisensor Data and Cross-Validation Technique for Merging Temporal Images for the Agricultural Performance Monitoring System

AutoSR4EO: An AutoML Approach to Super-Resolution for Earth Observation Images

Super-resolution (SR), a technique to increase the resolution of images, is a pre-processing step in the pipelines of applications of Earth observation (EO) data. The manual design and optimisation of SR models that are specific to every possible EO use case is a laborious process that creates a bottleneck for EO analysis. In this work, we develop an automated machine learning (AutoML) method to automate the creation of dataset-specific SR models. AutoML is the study of the automatic design of high-performance machine learning models. We present the following contributions. (i) We propose AutoSR4EO, an AutoML method for automatically constructing neural networks for SR. We design a search space based on state-of-the-art residual neural networks for SR and incorporate transfer learning. Our search space is extendable, making it possible to adapt AutoSR4EO to future developments in the field. (ii) We introduce a new real-world single-image SR (SISR) dataset, called SENT-NICFI. (iii) We evaluate the performance of AutoSR4EO on four different datasets against the performance of four state-of-the-art baselines and a vanilla AutoML SR method, with AutoSR4EO achieving the highest average ranking. Our results show that AutoSR4EO performs consistently well over all datasets, demonstrating that AutoML is a promising method for improving SR techniques for EO images.

Secure most significant bit plane compression based reversible data hiding in encrypted image technique using Huffman Ciphersystem

Secure most significant bit plane compression based reversible data hiding in encrypted image technique using Huffman Ciphersystem

FaceSigns: Semi-Fragile Watermarks for Media Authentication

Manipulated media is becoming a prominent threat due to the recent advances in realistic image and video synthesis techniques. There have been several attempts at detecting synthetically tampered media using machine learning classifiers. However, such classifiers do not generalize well to black-box image synthesis techniques and have been shown to be vulnerable to adversarial examples. To address these challenges, we introduce FaceSigns — a deep learning based semi-fragile watermarking technique that allows media authentication by verifying an invisible secret message embedded in the image pixels. Instead of identifying and detecting manipulated media using visual artifacts, we propose to proactively embed a semi-fragile watermark into a real image or video so that we can prove its authenticity when needed. FaceSigns is designed to be fragile to malicious manipulations or tampering while being robust to benign operations such as image/video compression, scaling, saturation, contrast adjustments etc. This allows images and videos shared over the internet to retain the verifiable watermark as long as a malicious modification technique is not applied. We demonstrate that our framework can embed a 128 bit secret as an imperceptible image watermark that can be recovered with a high bit recovery accuracy at several compression levels, while being non-recoverable when unseen malicious manipulations are applied. For a set of unseen benign and malicious manipulations studied in our work, our framework can reliably detect manipulated content with an AUC score of 0.996 which is significantly higher than prior image watermarking and steganography techniques.