Digital Image Research Articles

Impact of Surface Finishing on Ti6Al4V Voronoi Additively Manufactured Structures: Morphology, Dimensional Deviation, and Mechanical Behavior

Additively manufactured medical devices require proper surface finishing before their use to remove partially adhered particles and provide adequate surface roughness. The literature widely investigates regular lattice structures—mainly scaffolds with small pores to enhance osseointegration; however, only a few studies have addressed the impact of surface finishing on the dimensional deviation and the global and local mechanical responses of lattice samples. Therefore, the current research investigates the impact of biomedical surface finishing (i.e., corundum sandblasting and zirconia sandblasting) on Voronoi lattice structures produced by laser powder bed fusion (LPBF) with large pores and different thicknesses on the surface morphology and global and local mechanical behaviors. MicroCT and SEM are performed for the assessment of dimensional mismatch and surface evaluation. The mechanical properties are investigated with 2D digital image correlation (DIC) in quasi-static compression tests to estimate the impact of surface finishes on local maps of strain. In the quasi-static tests, both the global mechanical performances, as expected, and local 2D DIC strain maps were mainly affected by the strut thickness, and the impact of different surface finishings was irrelevant; on the contrary, different surface finishing processes led to differences in the dimensional deviation depending on the strut thickness. These results are relevant for designing lattice structures with thin struts that are integrated into medical prostheses that undergo AM.

WSSS-CRAM: precise segmentation of histopathological images via class region activation mapping

IntroductionFast, accurate, and automatic analysis of histopathological images using digital image processing and deep learning technology is a necessary task. Conventional histopathological image analysis algorithms require the manual design of features, while deep learning methods can achieve fast prediction and accurate analysis, but rely on the drive of a large amount of labeled data.MethodsIn this work, we introduce WSSS-CRAM a weakly-supervised semantic segmentation that can obtain detailed pixel-level labels from image-level annotated data. Specifically, we use a discriminative activation strategy to generate category-specific image activation maps via class labels. The category-specific activation maps are then post-processed using conditional random fields to obtain reliable regions that are directly used as ground-truth labels for the segmentation branch. Critically, the two steps of the pseudo-label acquisition and training segmentation model are integrated into an end-to-end model for joint training in this method.ResultsThrough quantitative evaluation and visualization results, we demonstrate that the framework can predict pixel-level labels from image-level labels, and also perform well when testing images without image-level annotations.DiscussionFuture, we consider extending the algorithm to different pathological datasets and types of tissue images to validate its generalization capability.

Design of optical performance for self-luminous pavement materials

The lower visibility in the nighttime environment makes higher accident risks during driving. Using self-luminous materials in pavement construction can effectively improve the luminance of load surface, thus significantly improve nighttime driving safety. This study has designed three types of self-luminous pavement materials (SLPM) with different optical properties by incorporating long-persistent luminescent aggregates into the mixture and using polymer binders with varying degrees of light transmittance, such as transparent polyurethane, transparent asphalt, and traditional asphalt. On this basis, utilizing digital image processing techniques to compare the optical performance differences among these three SLPM and established corresponding computational models based on the obtained data to enable efficient and accurate evaluation of the optical properties of pavement materials. The experimental results indicate that the durability performance of SLPM gradually diminishes over time, and this degree of attenuation is negatively correlated with the light transmittance of the binder. Specifically, a higher light transmittance of the binder results in a smaller decline in SLPM durability, whereas a lower transmittance leads to a more significant reduction. The optical performance calculation model established by this research can predict the reflective performance of SLPM through the optical parameters of aggregate, binders, and the doping amount of long persistent luminescent materials. In addition, the model can accurately calculate the minimum content of self-luminous materials according to the optical requirements required by the project, which can provide a reference for the performance evaluation of self-luminous pavement materials and the selection of optimal proportioning scheme.

Experimental study on laser-assisted machining of SiCp/Al composites considering laser preheating history based on in-situ observation

The application of laser-assisted machining (LAM) has been demonstrated to improve the surface integrity of SiCp/Al composites. The dynamic process of laser preheating affects the deformation mechanism of materials. In this paper, the laser preheating history under different parameters is characterized by the changes in peak temperature and instantaneous preheating temperature. SiCp/Al cutting experiments under different laser preheating histories are carried out based on the high-speed imaging system. The influence of laser preheating history on the shear angle and the spatial–temporal distribution of the physical information in the shear zone are analyzed by the digital image correlation (DIC) method. The deformation characteristics during the SiCp/Al composites cutting process under different laser preheating histories are discussed. Additionally, the influence of the laser preheating history on the surface integrity is analyzed according to the changes of the Al matrix and SiC particles under different combinations of peak temperature and instantaneous preheating temperature. It has been demonstrated that the deformation degree in the shear zone during LAM of SiCp/Al composites is depended on the material softening degree related to the peak temperature and the material deformation temperature dominated by the instantaneous preheating temperature. The deformation uniformity is primarily determined by the deformation temperature. An appropriate material softening degree and deformation temperature can improve the surface integrity. The results provide a foundation for further mechanism studies for cutting load and surface integrity in LAM of SiCp/Al composites.

Mitigating Adversarial Attacks in Object Detection through Conditional Diffusion Models

The field of object detection has witnessed significant advancements in recent years, thanks to the remarkable progress in artificial intelligence and deep learning. These breakthroughs have significantly enhanced the accuracy and efficiency of detecting and categorizing objects in digital images. Nonetheless, contemporary object detection technologies have certain limitations, such as their inability to counter white-box attacks, insufficient denoising, suboptimal reconstruction, and gradient confusion. To overcome these hurdles, this study proposes an innovative approach that uses conditional diffusion models to perturb adversarial examples. The process begins with the application of a random chessboard mask to the adversarial example, followed by the addition of a slight noise to fill the masked area during the forward process. The adversarial image is then restored to its original form through a reverse generative process that only considers the masked pixels, not the entire image. Next, we use the complement of the initial mask as the mask for the second stage to reconstruct the image once more. This two-stage masking process allows for the complete removal of global disturbances and aids in image reconstruction. In particular, we employ a conditional diffusion model based on a class-conditional U-Net architecture, with the source image further conditioned through concatenation. Our method outperforms the recently introduced HARP method by 5% and 6.5% in mAP on the COCO2017 and PASCAL VOC datasets, respectively, under non-APT PGD attacks. Comprehensive experimental results confirm that our method can effectively restore adversarial examples, demonstrating its practical utility.

Analysis of virtual simulation training teaching problems of intellectual property law based on real working situations

This article aims to take AI intelligent training software and real business drills of intellectual property as the training mode to carry out intellectual property-related training in fields such as data intellectual property creation, application, management, transformation, and education, and truly achieve the educational goal of integrating science, education and industry. First, innovate the intellectual property training platform. Through AI question-and-answer led by digital human images, introduce front-line practical experts in the field of intellectual property into the virtual simulation training teaching process, build a training scene for the entire chain protection of intellectual property and real working situations, and improve the practical effect of training and cultivating intellectual property talents. Second, deepen the social service system of intellectual property. Project team members guide students and trainees to actively participate in patent navigation, patent group projects, pilot projects, intellectual property protection projects, etc., exercise and promote the continuous improvement of the social service capabilities of various intellectual property talents. At the end of the article, it is proposed that while intellectual property law teaching adapts to the development needs of the digital age, it should take the construction and teaching application of a virtual simulation experiment platform in real work situations as the core, respond to the practical needs of social practice work, and improve students' ability to adapt to society. The organic integration of AI technology, digital human technology and practical training teaching of intellectual property law has certain reference significance for the realization of the metaverse vision.

B-006 Use of Sysmex UN-3000 Research Parameters for Semiquantitative Analysis of Urine Microscopic Elements

Abstract Background Microscopic examination of urine is an important component of urine analysis. However, it is labor-intensive and requires a skilled microscopist. The Sysmex UN-3000 (Kobe, JP) combines a Siemens CLINITEK Novus (Munich, GER) with a Sysmex UF-5000 flow cytometer and Sysmex UD-10 digital imaging device for fully automated urine analysis. It is FDA-approved for quantitation of red and white blood cells, epithelial cells, casts, and bacteria. It also flags for other particles such as sperm, yeast, pathologic casts, and crystals, which can then be confirmed on the UD-10. We sought to determine if and how the UF-5000 research parameter results could be used for semi-quantitative grading of pathologic casts, squamous epithelial cells, non-squamous epithelial cells, and crystals to match our reporting system. Methods Urine samples were analyzed by the UN-3000 and by manual review of urine sediment per laboratory protocols. Epithelial cells and crystals grades were none seen, few, moderate, many. Cast grades were 0, 1-3, 4-10, and >10. Results of manual microscopy were used to determine numerical cutoffs for the UF-5000 results which would agree within one grade in ≥90% of samples. Sperm and yeast are reported as either present or absent. For these particles, manual microscopy was compared to the results of the UF-5000 flagging and subsequent review of the UD-10 images. Results Rate of agreement between the methods is shown in the table. For sperm the UN-3000 showed 91.7% sensitivity and 100% specificity; for yeast, 93.8% sensitivity and 97.7% specificity. Conclusions We were able to use the research parameters from the UF-5000, in conjunction with the UD-10, to perform semi-quantitative grading. Adoption of this analyzer increases automation, with 70% of samples auto-validating per our protocols, while still conforming to our previous reporting system.

A Zero-Watermarking Algorithm Based on Vortex-like Texture Feature Descriptors

For effective copyright protection of digital images, this paper proposes a zero-watermarking algorithm based on local image feature information. The feature matrix of the algorithm is derived from the keypoint set determined by the Speeded-Up Robust Features (SURF) algorithm, and it calculates both the gradient feature descriptors and the vortex-like texture feature (VTF) descriptors of the keypoint set. Unlike traditional texture feature descriptors, the vortex-like texture feature descriptors proposed in this paper contain richer information and exhibit better stability. The advantage of this algorithm lies in its ability to calculate the keypoints of the digital image and provide a stable vector description of the local features of these keypoints, thereby reducing the amount of erroneous information introduced during attacks. Analysis of experimental data shows that the algorithm has good effectiveness, distinguishability, security, and robustness.

Assessing generalizability of an AI-based visual test for cervical cancer screening.

A number of challenges hinder artificial intelligence (AI) models from effective clinical translation. Foremost among these challenges is the lack of generalizability, which is defined as the ability of a model to perform well on datasets that have different characteristics from the training data. We recently investigated the development of an AI pipeline on digital images of the cervix, utilizing a multi-heterogeneous dataset of 9,462 women (17,013 images) and a multi-stage model selection and optimization approach, to generate a diagnostic classifier able to classify images of the cervix into "normal", "indeterminate" and "precancer/cancer" (denoted as "precancer+") categories. In this work, we investigate the performance of this multiclass classifier on external data not utilized in training and internal validation, to assess the generalizability of the classifier when moving to new settings. We assessed both the classification performance and repeatability of our classifier model across the two axes of heterogeneity present in our dataset: image capture device and geography, utilizing both out-of-the-box inference and retraining with external data. Our results demonstrate that device-level heterogeneity affects our model performance more than geography-level heterogeneity. Classification performance of our model is strong on images from a new geography without retraining, while incremental retraining with inclusion of images from a new device progressively improves classification performance on that device up to a point of saturation. Repeatability of our model is relatively unaffected by data heterogeneity and remains strong throughout. Our work supports the need for optimized retraining approaches that address data heterogeneity (e.g., when moving to a new device) to facilitate effective use of AI models in new settings.

A Digital Image Method for Calculating the Working Chamber Volume of a Combined Profile Scroll Compressor

Background: The efficient and accurate calculation of working chamber volume can greatly contribute to the optimized design efficiency of the combined profile scroll compressor, but current papers and patents lack research on the method of calculating the working chamber volume of a combined profile scroll compressor. Objective: A method of using digital image processing is proposed to efficiently calculate the working chamber volume of a combined scroll compressor. Methods: This method discretizes and reconstructs the scroll profile that forms the working chamber into a coordinate sequence with equal involute angle intervals. By calculating the relationship between the coordinate sequence and the rotation angle, a general coordinate sequence that forms the working chamber profile is obtained. The real-time changes in the projection image of the working chamber during all the suction, compression and discharge processes can be accurately depicted. The digital image is processed to obtain the actual projection area of the working chamber, and thus, the volume of the working chamber is accurately calculated. Results: The digital image method can accurately calculate the volume change of the working chamber of the combined profile scroll compressor by selecting the appropriate involute angle interval and digital image size, with a mean relative error of less than 1%. At the beginning of suction and the end of discharge, the calculated volume has been found to have poor accuracy, with a maximum relative error higher than 10%. Conclusion: The digital image method has been found to have high accuracy, greatly reduce the difficulty of the analysis of the working chamber volume, and promote the design optimization of the combined profile scroll compressor, thus broadening the idea for the calculation method of the working chamber volume of the scroll compressor.

A hardware architecture for single and multiple ellipse detection using genetic algorithms and high-level synthesis tools

Ellipse detection techniques are often developed and validated in software environments, neglecting the critical consideration of computational efficiency and resource constraints prevalent in embedded systems. Furthermore, programmable logic devices, notably Field Programmable Gate Arrays (FPGAs), have emerged as indispensable assets for enhancing performance and expediting various processing applications. In the realm of computational efficiency, hardware implementations have the flexibility to tailor the required arithmetic for various applications using fixed-point representation. This approach enables faster computations while upholding adequate accuracy, resulting in reduced resource and energy consumption compared to software applications that rely on higher clock speeds, which often lead to increased resource and energy consumption. Additionally, hardware solutions provide portability and are suitable for resource-constrained and battery-powered applications. This study introduces a novel hardware architecture in the form of an intellectual property core that harnesses the capabilities of a genetic algorithm to detect single and multi ellipses in digital images. In general, genetic algorithms have been demonstrated to be an alternative that shows better results than those based on traditional methods such as the Hough Transform and Random Sample Consensus, particularly in terms of accuracy, flexibility, and robustness. Our genetic algorithm randomly takes five edge points as parameters from the image tested, creating an individual treated as a potential candidate ellipse. The fitness evaluation function determines whether the candidate ellipse truly exists in the image space. The core is designed using Vitis High-Level Synthesis (HLS), a powerful tool that converts C or C＋＋functions into Register-Transfer Level (RTL) code, including VHDL and Verilog. The implementation and testing of the ellipse detection system were carried out on the PYNQ-Z1, a cost-effective development board housing the Xilinx Zynq-7000 System-on-Chip (SoC). PYNQ, an open-source framework, seamlessly integrates programmable logic with a dual-core ARM Cortex-A9 processor, offering the flexibility of Python programming for the onboard SoC processor. The experimental results, based on synthetic and real images, some of them with the presence of noise processed by the developed ellipse detection system, highlight the intellectual property core’s exceptional suitability for resource-constrained embedded systems. Notably, it achieves remarkable performance and accuracy rates, consistently exceeding 99% in most cases. This research aims to contribute to the advancement of hardware-accelerated ellipse detection, catering to the demanding requirements of real-time applications while minimizing resource consumption.

Measures of angularity in digital images.

In light of the growing interest in studying the affective and aesthetic attributes of curvature, the present paper describes four digital image processing techniques that can be used to objectively discriminate between angular and curvilinear stimuli. MATLAB scripts for each of the techniques accompany the paper. Three studies are then reported that evaluate the efficacy of five metrics, derived from the four techniques, at quantifying the degree of angularity depicted in an image. Images of simple polygons (Study 1), artistic drawings of everyday objects (Study 2), and real-world objects, typefaces, and abstract patterns (Study 3) were analyzed. Logistic regression models were used to determine the relative importance of the metrics at distinguishing between angular and curvilinear items. With one exception, all of the metrics were capable of distinguishing between angular and curvilinear items at a level above chance, but some metrics were better at doing so than others, and their discriminative capacity was influenced by the characteristics of the image. The strengths and limitations of the metrics are discussed, as well as some practical recommendations.

Deep learning-based image encryption techniques: Fundamentals, current trends, challenges and future directions

In recent years, the number of digital images has grown exponentially because of the widespread use of fast internet and smart devices. The integrity authentication of these images is a major concern for the research community. So, the encryption schemes that are commonly used to protect these images are an important subject for many potential applications. This paper presents a comprehensive survey of recent image encryption techniques using deep learning models. First, we explain the reasons that image encryption using deep learning models is beneficial to researchers and the public. Second, we discuss various state-of-art encryption techniques using deep learning models and offer technical summaries of popular techniques. Third, we provide a comparative analysis of our survey and existing state-of-the-art surveys. Finally, by investigating existing deep learning-based encryption, we identify several important research challenges and possible solutions including standard security metrics. To the best of our knowledge, we are the first researchers to do a detailed survey of deep learning-based image encryption for digital images.

Radiographic Measurements of the Hip Joint in Capuchin Monkeys (Sapajus spp.).

Radiographs are useful for the initial evaluation of the hip joints. The information can be utilized for the betterment of animal health or other goals such as anatomic studies and gait analysis, among others. Therefore, this study aimed to evaluate radiographic measurements of the hip joint in capuchin monkeys, kept under human care at a reference center for wildlife. Twelve capuchin monkeys (Sapajus spp.) (three adult males, seven adult females, and two sub-adult females) were evaluated. Ventrodorsal radiographic views were taken under chemical restraint. All measurements on the digital images were performed in triplicate by one examiner. None of the measurements evaluated were statistically different between males and females. No statistical differences were found between hind limbs. The mean (±SD) Norberg angle was 104.92° (±2.82°) and the Wiberg angle was 15.26° (±1.86°). The percentage of the femoral head covered by the acetabulum was 68.57% (±3.65%) and the acetabular index depth to width ratio was 54.66% (±3.85%). In conclusion, the radiographic measurements showed certain morphological features of the hip joint in Sapajus spp. that contribute to improving species knowledge.

Deep-learning classification of teat-end conditions in Holstein cattle

As a means of preventing mastitis, deep learning for classifying teat-end conditions in dairy cows has not yet been optimized. By using 1426 digital images of dairy cow udders, the extent of teat-end hyperkeratosis was assessed using a four-point scale. Several deep-learning networks based on the transfer learning approach have been used to evaluate the conditions of the teat ends displayed in the digital images. The images of the teat ends were partitioned into training (70 %) and validation datasets (15 %); afterwards, the network was evaluated based on the remaining test dataset (15 %). The results demonstrated that eight different ImageNet models consistently achieved high accuracy (80.3–86.6 %). The areas under the receiver operating characteristic curves for the normal, smooth, rough, and very rough classification scores in the test data set ranged from 0.825 to 0.999. Thus, improved accuracy in image-based classification of teat tissue conditions in dairy cattle using deep learning requires more training images. This method could help farmers reduce the risks of intramammary infections, decrease the use of antimicrobials, and better manage costs associated with mastitis detection and treatment.

H-partial uniform spaces and their application in the compression of digital images

Fixed point theorem is very important tool in different branches of mathematics. In this paper, we introduce partial uniform spaces as a generalization of uniform spaces and metric spaces; and study some basic properties. Various examples support the theory. We prove fixed point theorems for H-partial uniform spaces, by using a mapping called E-distance function. Finally, we give the applications of these fixed point theorems to compress digital images.

Characterization of Coal Permeability Considering Fracture Shape — Using the MP-Otsu Threshold Segmentation Algorithm

Coal seepage simulation based on digital images is capable of visualizing the three-dimensional seepage process and providing permeability data. However, Otsu's conventional thresholding algorithm is prone to image mis-segmentation for extracting pores and fractures, which makes the reconstructed model unable to reestablish the coal structure truly. The present work first explains the failure mechanism of the conventional Otsu algorithm for image segmentation based on computed tomography (CT) images of coal. Three weighting factors, including the slope of the mineral content fitting curve against the Otsu threshold and the approximate proportion of the target in the image, are introduced to enhance the accuracy of the threshold relation. In continuing, the MP-Otsu thresholding algorithm is proposed. Finally, the 3D fracture model of the coal is reconstructed from the CT image via the MP-Otsu thresholding algorithm. Focusing on the shape of the model, the permeability is then calculated on the basis of the pipe flow model and the parallel plate model. The contribution of fractures with various radii in the pipe model and different apertures in the parallel plate model is appropriately incorporated into the permeability evaluation. The obtained results reveal that the presence of mineral components and the low porosity of coal are the chief reasons for the failure of the conventional Otsu algorithm in the digital image segmentation of coal. The porosities calculated by the improved MP-Otsu algorithm are predicted to be 7.13%-72.03% lower than those of the Otsu’s method, which effectively improves the over-segmentation of fractures by Otsu and could accurately extract fractures in images. The permeability model, considering the fracture shape, gives results similar to the simulation results, and parallel plate fractures remarkably affect the permeability of the coal body. In particular, fractures with radius R > 7 μm and apertures b > 50 μm account for more than 90% of the permeability and thus play a key role in the seepage process.

Predictive Analysis of Breast Cancer from Full-Field Digital Mammography Images using Residual Network

Breast cancer has been a significant contributor to cancer-related mortality, but advancements in early detection through regular mammography and improvements in treatment modalities have contributed to declining mortality rates in several regions. This study presents a novel approach to cancer diagnosis utilizing Full-Field Digital Mammography images through predictive analysis methods. By using predictive analytic techniques and mammography images, this study offers a novel way to cancer detection. The research involves the application of deep learning techniques to extract valuable insights from cancer images captured by mammography devices. The CBIS-DDSM (Curated Breast Imaging Subset of Digital Database for Screening Mammography) dataset including images from patients with varying types and stages of cancer, is collected and pre-processed to ensure uniformity and quality. Relevant features, including color, texture, and shape characteristics, are extracted, and a rigorous feature selection process is employed to identify discriminative markers. The Residual Network (ResNet) model is selected and trained on the dataset, with a focus on classification accuracy and robust predictive performance. Validation metrics, such as accuracy, IoU (Intersection over Union) score, dice score, and ROC (Receiver Operating Characteristic) curve are employed to evaluate the model’s efficiency. After analysis, the proposed method had the best degree of mass lesion detection accuracy, at 99.24%. This research contributes to the advancement of non-invasive and efficient diagnostic tools, potentially enhancing early detection and intervention in cancer patients. The proposed method not only demonstrates promising results in terms of diagnostic accuracy but also emphasizes interpretability, seamless integration into clinical workflows, and adherence to ethical standards.

Mandibular asymmetry. Hemimandibular hyperplasia vs. hemimandiubular elongation

Introduction. Hemimandibular hyperplasia and elongation create a developmental asymmetry characterized by volumetric changes of the hemimandible. Both entities cause progressive aesthetic and functional problems causing depression with poor quality of life. The treatment of these anomalies is combined, orthodontic and surgical Material and methods. The study was based on the analysis of bibliographic sources as well as the rehabilitation of two patients diagnosed with hemimandibular hyperplasia and hemimandibular elongation. Results. The specialized literature proposes a variety of surgical techniques for osteotomy of the mandible to correct mandibular asymmetries. The surgical planning of the interventions was carried out using the digital protocol, with the manufacture of intermediate and final splints. Conclusions. Contemporary digital medical imaging in combination with virtual surgical planning allows a patient specific treatment protocol, a fact that significantly contributes to optimizing not only the treatment plan but also the surgery itself.

Disappearance of the face: From early photography to facial recognition systems

In this article I analyse the hidden genealogical link between portrait photography, used for criminological and psychiatric purposes, and contemporary systems of biometric identification of the human face. The aim is to highlight the shift between the emphasis on the importance of the human ‘expert eye’ in recognizing the face when talking about nineteenth-century photography and the use of computer technology that produces and reads digital facial images. In both cases, however, these are modes and variants of reducing and flattening the human face; the face itself disappears under the onslaught of technologies of vision and mediation, becoming a mere data set. Special attention is devoted to the pose, the frontal view, which is the technological a priori of the empirical possibility of recognition, articulated through the history of visualization of the face from early portrait photography to facial recognition systems. Consequently, what we call the ‘face’ is a simulacra of individual the face presented to the apparatus: the ‘real’ face is transposed upon multiple technological layers (camera – plates – photographic surface), losing its characteristic features to be re-written according to specific techniques of measurement.