Color Distribution Research Articles

Microplastic contamination of packaged spirulina products.

Microplastic (MP) contamination in commercially soldspirulina products has not been previously investigated. In this study, 29 spirulina samples in various packaging types were purchased from different brands and origins to assess the presence of MPs. Microplastic analysis was conducted using microscopic and μ-Raman techniques. To ascertain whether the content is indeed spirulina and make a comparison with the MP level, C-Phycocyanin levels were also analyzed. A total of 251 MP-like particles were observed. Out of the 29 examined packaged spirulina brands, 26 showed potential MPs upon visual inspection, with 35 particles confirmed as MPs (73% of the analyzed particles). The mean abundance of MPs was estimated at 13.77 ± 2.45 MPs/100g dw. Powdered spirulina had a higher but not statistically significant MP abundance (17.34 ± 4.22 MPs/100g dw) compared to capsule/tablet forms (10.43 ± 2.45 MPs/100g dw). Fragments accounted for 38.3% while fibers constituted 61.7% of the identified MPs, with sizes ranging from 0.07 to 2.15mm for fragments and 0.19 to 5.691mm for fibers. The color distribution of MPs in spirulina samples was predominantly blue (52.8%), followed by black (25.4%), white (10.9%), and others (10.9%). Ten synthetic polymers and cellulose were identified through μ-Raman analysis, with polypropylene (31.6%) and polystyrene (8.3%) being the most prevalent. The correlation between C-Phycocyanin and MPs concentrations, was not found statistically significant. The abundance and composition of MPs were found to be influenced by packaging and processing stages. Identifying potential sources of MPs in spirulina products and evaluating their risks to human health is crucial.

Analyzing the color availability of AI‐generated posters based on K‐means clustering: 74% orange, 38% cyan, 32% yellow, and 28% blue‐cyan

AbstractIn this exploratory study, we delved deeply into the intricate interplay of color choices within AI‐generated and human‐designed posters, analyzing a sample of 120 instances from each category. While it is suggested that human designers may integrate cultural, emotional, and situational contexts into their creations, AI models largely base their selections on vast datasets and pattern recognition. Although AI exhibited prowess in replicating established design parameters, the study underlined the importance of critically assessing its outputs. The quantitative analysis illuminated overarching similarities in primary color selections. However, the AI's diversity in color remains less concentrated than that of human, suggesting a gap in the AI's capacity to match human expertise in color proportioning and distribution. As AI continues to evolve, it is crucial to discern its capabilities and potential limitations in the design domain, ensuring it augments human creativity rather than supplanting it. Notably, the research refrains from seeking human validation, aiming instead for an objective, data‐driven reflection on the convergences and divergences between AI‐generated and human designs.

BCD-net: Stain separation of histological images using deep variational Bayesian blind color deconvolution

Histological images are often tainted with two or more stains to reveal their underlying structures. Blind Color Deconvolution (BCD) techniques separate colors (stains) and structural information (concentrations), which is useful for the processing, data augmentation, and classification of such images.Classical analytical BCD methods are typically computationally expensive in two distinct ways. First, estimating the colors and concentrations corresponding to a given image is a time-consuming process. Second, the entire estimation procedure must be performed independently for each image.In contrast, Deep Learning (DL) methods involve high training costs, but once trained, they are able to directly process unseen images. The application of DL to BCD has been limited by the absence of extensive databases containing ground truth color and concentrations. In this work, we propose BCD-Net, a deep variational Bayesian neural network for stain separation and concentration estimation. Under this framework, we address the challenge of lacking ground truth data by leveraging Bayesian modeling and inference techniques.We propose to use a prior distribution on the stain colors, and a simple flat prior on the concentrations. BCD-Net is trained by maximizing the evidence lower bound of the observed images. The loss function comprises two essential components: fidelity to the observed images and the Kullback-Leibler divergence between the estimated posterior distribution of colors and the selected prior.The model is trained, validated, and tested on two multicenter databases: Camelyon-17 and Warwick stain separation benchmark. The proposed approach is tested on image reconstruction, stain separation, and cancer classification. It performs well when contrasted with classical non-amortized methods and offers a substantial computational time advantage. This marks a significant step forward in the application of DL techniques to address BCD and paves the way for new approaches.

A colorimetric sensing platform with smartphone for organophosphorus pesticides detection based on PANI-MnO2 nanozyme

Organophosphorus pesticides (OPs) are of great concern due to its potential harms on human health and the environment. Herein, a budget-friendly, rapid and convenient colorimetric sensing platform is developed for detection of OPs in the environmental and food samples. The sensing element, PANI-MnO2 nanozyme with excellent oxidase mimetic activity is synthesized at room temperature, which is able to directly oxidize 3,3,5,5-tetramethylbenzidine (TMB) to generate blue colored oxidized TMB (OxTMB) within 2 min. Ascorbic acid (AA) can inhibit the oxidization reaction of TMB, consequently causing the blue color fading. Ascorbic acid 2-phosphate (AAP) could be hydrolyzed to produce AA by alkaline phosphatase (ALP). In the presence of OPs can effectively decrease ALP activity, resulting in the recovery of catalytic activity of PANI-MnO2. Therefore, sensitive and selective OPs detection is achieved. Under the optimal conditions, excellent detection performance in term of glyphosate as a model is achieved with a linear range from 0.50 to 50 μM, the detection limit is 0.39 μM (S/N = 3). The utility of method is further improved by combining a portable smartphone platform with a color picking application. The colorimetric platform achieves instrument-free detection of OPs and overcomes the uneven color distribution of traditional paper-based chip, providing an alternative strategy for the qualitative discernment and semi-quantitative analysis of OPs on-site.

In situ visualization of misorientation-dependent hydrogen diffusion at grain boundaries of pure polycrystalline Ni using a hydrogen video imaging system

Hydrogen, the key element for solving the major environmental issues, exhibits rapid diffusion and localization in lattice defects of metals, which leads to embrittlement. However, the microstructure-hydrogen interactions in metallic materials are poorly understood. Therefore, a technique for capturing the microstructure-dependent hydrogen diffusion in real-time has long been targeted by researchers. Here, we successfully visualized the preferential hydrogen diffusion at grain boundaries (GBs) of pure Ni using a hydrogen video imaging system. A thin polyaniline layer was utilized as a hydrogenochromic sensor, and the hydrogen distribution was analyzed based on the color distribution of polyaniline. The system revealed the misorientation-dependent hydrogen flux at GBs of pure polycrystalline Ni with a micrometer-scale spatial resolution. It suggests that the geometrical structure of the GBs is a critical factor for preferential hydrogen diffusion. The use of the hydrogen video imaging system further advances the mechanistic understanding of hydrogen-material atomic interactions in polycrystalline metals and critically facilitates the development of hydrogen-related materials.

Assessing the utility of high spectral resolution lidar for measuring particulate backscatter in the ocean and evaluating satellite ocean color retrievals

Airborne high spectral resolution lidar (HSRL) measurements of ocean particulate backscatter (bbp) offer dramatic improvements in spatiotemporal coverage over in situ techniques, filling observational “blind spots” that limit our ability to study ocean processes. However, the technique has been assessed in only a few limited cases, and uncertainties remain regarding its applicability across a diversity of optical domains. In this study, we present the first comprehensive comparison of bbp derived from airborne HSRL, satellite ocean color, and in situ measurements to increase confidence in HSRL bbp retrievals and to demonstrate the value of airborne HSRL for assessing/improving satellite ocean color algorithms. Retrievals of bbp performed using the NASA Langley HSRL-1 instrument agreed with in situ measurements performed across a diversity of optical and ecological domains. Comparisons across multiple campaigns revealed regional and seasonal dependencies in the ocean color retrievals that likely resulted from applying a single configuration of the ocean color retrieval across multiple distinct domains. In two case studies, atmospheric measurements from HSRL-1 and systematic differences in bbp between HSRL and the 2018 NASA ocean color distribution provided evidence of insufficient atmospheric correction of the ocean color retrieval. These differences in bbp were absent from comparisons against the 2022 ocean color distribution, suggesting that changes made to the algorithm resulted in improved retrievals. These cases highlight the advantages of airborne HSRL for assessing and improving ocean color retrievals, namely its ability to provide simultaneous, independent profiles of atmosphere and ocean optical properties, and improvements in the spatiotemporal coverage of satellite matchups.

Clustering and Segmentation of Adhesive Pests in Apple Orchards Based on GMM-DC

The segmentation of individual pests is a prerequisite for pest feature extraction and identification. To address the issue of pest adhesion in the apple orchard pest identification process, this research proposed a pest adhesion image segmentation method based on Gaussian Mixture Model with Density and Curvature Weighting (GMM-DC). First, in the HSV color space, an image was desaturated by adjusting the hue and inverting to mitigate threshold crossing points. Subsequently, threshold segmentation and contour selection methods were used to separate the image background. Next, a shape factor was introduced to determine the regions and quantities of adhering pests, thereby determining the number of model clustering clusters. Then, point cloud reconstruction was performed based on the color and spatial distribution features of the pests. To construct the GMM-DC segmentation model, a spatial density (SD) and spatial curvature (SC) information function were designed and embedded in the GMM. Finally, experimental analysis was conducted on the collected apple orchard pest images. The results showed that GMM-DC achieved an average accurate segmentation rate of 95.75%, an average over-segmentation rate of 2.83%, and an average under-segmentation rate of 1.42%. These results significantly outperformed traditional image segmentation methods. In addition, the original and improved Mask R-CNN models were used as recognition models, and the mean Average Precision was used as the evaluation metric. Recognition experiments were conducted on pest images with and without the proposed method. The results show the mean Average Precision for pest images segmented with the proposed method as 92.43% and 96.75%. This indicates an improvement of 13.01% and 12.18% in average recognition accuracy, respectively. The experimental results demonstrate that this method provides a theoretical and methodological foundation for accurate pest identification in orchards.

ECR Spotlight – Rachel Thayer

ECR Spotlight is a series of interviews with early-career authors from a selection of papers published in Journal of Experimental Biology and aims to promote not only the diversity of early-career researchers (ECRs) working in experimental biology during our centenary year but also the huge variety of animals and physiological systems that are essential for the ‘comparative’ approach. Rachel Thayer is an author on ‘ A meta-analysis of butterfly structural colors: their color range, distribution and biological production’, published in JEB. Rachel conducted the research described in this article while a PhD student in Nipam Patel's lab at University of California Berkeley/Marine Biological Laboratory, USA. She is now a Postdoctoral Fellow in the lab of David Begun at University of California, Davis, USA, investigating the genetic basis of complex, adaptive traits, evolutionary innovation and the relationship between microevolutionary and macroevolutionary processes and their diversity.

A meta-analysis of butterfly structural colors: their color range, distribution and biological production.

Butterfly scales are among the richest natural sources of optical nanostructures, which produce structural color and iridescence. Several recurring nanostructure types have been described, such as ridge multilayers, gyroids and lower lamina thin films. While the optical mechanisms of these nanostructure classes are known, their phylogenetic distributions and functional ranges have not been described in detail. In this Review, we examine a century of research on the biological production of structural colors, including their evolution, development and genetic regulation. We have also created a database of more than 300 optical nanostructures in butterflies and conducted a meta-analysis of the color range, abundance and phylogenetic distribution of each nanostructure class. Butterfly structural colors are ubiquitous in short wavelengths but extremely rare in long wavelengths, especially red. In particular, blue wavelengths (around 450 nm) occur in more clades and are produced by more kinds of nanostructures than other hues. Nanostructure categories differ in prevalence, phylogenetic distribution, color range and brightness. For example, lamina thin films are the least bright; perforated lumen multilayers occur most often but are almost entirely restricted to the family Lycaenidae; and 3D photonic crystals, including gyroids, have the narrowest wavelength range (from about 450 to 550 nm). We discuss the implications of these patterns in terms of nanostructure evolution, physical constraint and relationships to pigmentary color. Finally, we highlight opportunities for future research, such as analyses of subadult and Hesperid structural colors and the identification of genes that directly build the nanostructures, with relevance for biomimetic engineering.

Spectroscopic Links among Giant Planet Irregular Satellites and Trojans

We collect near-infrared spectra (∼0.75–2.55 μm) of four Jovian irregular satellites and visible spectra (∼0.32–1.00 μm) of two Jovian irregular satellites, two Uranian irregular satellites, and four Neptune Trojans. We find close similarities between the observed Jovian irregular satellites and previously characterized Jovian Trojans. However, irregular satellites’ unique collisional histories complicate comparisons to other groups. Laboratory study of CM and CI chondrites shows that grain size and regolith packing conditions strongly affect spectra of dark, carbonaceous materials. We hypothesize that different activity histories of these objects, which may have originally contained volatile ices that subsequently sublimated, could cause differences in regolith grain size or packing properties and therefore drive spectral variation. The Uranian satellites Sycorax and Caliban appear similar to TNOs. However, we detect a feature near 0.7 μm on Sycorax, suggesting the presence of hydrated materials. While the sample of Neptune Trojans have more neutral spectra than the Uranian satellites we observe, they remain consistent with the broad color distribution of the Kuiper Belt. We detect a possible feature near 0.65–0.70 μm on Neptune Trojan 2006 RJ103, suggesting that hydrated material may also be present in this population. Characterizing hydrated materials in the outer solar system may provide critical context regarding the origins of hydrated CI and CM chondrite meteorites. We discuss how the hydration state(s) of the irregular satellites constrains the thermal histories of the interiors of their parent bodies, which may have formed among the primordial Kuiper Belt.

A preliminary investigation of niche clothing brand design optimization based on ADDIE model

Abstract This paper utilizes the ADDIE model and feature extraction method to analyze and study the design of niche clothing brands. Through the study of the ADDIE model, the model’s components are applied to the design of niche clothing brands. The three elements affecting the design of clothing brands are studied, and the main reasons affecting niche clothing brands are explored through the analysis of the integrity, normality and planning of clothing design. Combining color feature extraction methods to calculate moments is used to describe the color distribution of the images. The results show that the niche clothing brand design has great uniqueness compared to the mass clothing brand design, with an embodiment rate of 75%. Compared to other feature extraction methods, the detection rate for color feature extraction is 65%, corner point detection is 70%, and speckle detection is 59%.

The Color Concept of Baju Bodo as Traditional Clothing at Schools in Makassar: Philosophical Perspective and Ethno-Policy Framework

This study aims to examine the philosophical perspective of the Bodo shirt as one of the traditional clothes of South Sulawesi, which will be applied in the regulations for the use of traditional clothes in schools in Makassar City. This is based on the Minister of Education and Culture Regulation Number 50 of 2022 concerning School Uniforms for Students at the Elementary and Secondary Education Levels. This study used a qualitative approach with an ethnomethodological research design. Data analysis used indexicality analysis and reflexivity analysis. This analysis is intended to provide an overview of group agreement regarding the use of Baju Bodo in elementary school policies as one of the local cultural assets as well as the implementation of ethno-policies for cultural literacy. Determination of informants was carried out by means of purposive sampling, namely researchers who determine the informants themselves to be interviewed based on representative considerations. The informants in this study included teachers, school principals, and community leaders who knew about the philosophy of the Baju Bodo and the color schemes used in the use of Baju Bodo for the Bugis people of South Sulawesi. The results of the study show that in a philosophical study the structure of the baju bodo represents the characteristics of Bugis-Makassar women in general. The distribution of colors in Baju Bodo is divided into two types, the first is in everyday life, such as children under the age of 10 wearing Baju Bodo which are usually called Waju Pella-Pella, this shirt is Ivory Yellow. This shirt is called waju pella-pella or butterfly because it depicts the world of small children who are full of joy. The yellow color of ivory is an analogy for the child to mature quickly and be ready to face life's challenges. Derived from the word maridi (ivory yellow) which means ripe.

The lightness distribution of the red color for the nail plate related to human healthy level

The water content of the human nail plate may be a health indicator related to human beings. In Chinese traditional medical science, the traditional Chinese medicine physician usually examines the nail plate conditions combined with other body physical symptoms to determine the physical health level. Usually, for people with healthy bodies, the nail plates appear shining pink color. On the other hand, people get ills, their nail plates will appear in different colors except for shining pink color nail plates. The pink color nail plates have come from the red blood flowing through the body capillaries. A healthy people, the blood flow through capillaries should be very smooth. Obviously, the capillaries inside the human nail plates with a smooth flow of blood will cause the human nail plates to appear pink in color. Inspect the lightness distribution of the digital photo red color for the human nail plates only, a very clear result can be seen from the lightness distribution. The lightness distribution of the red color of the digital photo from the healthy human nail plate is wider than the lightness distribution of the digital photo from the dry nail plate.

Panchromatic Simulated Galaxy Observations from the NIHAO Project

We present simulated galaxy spectral energy distributions (SEDs) from the far-ultraviolet (FUV) through the far-infrared (FIR), created using hydrodynamic simulations and radiative transfer calculations, suitable for the validation of SED modeling techniques. SED modeling is an essential tool for inferring star formation histories from nearby galaxy observations, but it is fraught with difficulty due to our incomplete understanding of stellar populations, chemical enrichment processes, and the nonlinear, geometry-dependent effects of dust on our observations. Our simulated SEDs will allow us to assess the accuracy of these inferences against galaxies with known ground truth. To create the SEDs, we use simulated galaxies from the Numerical Investigation of Hundred Astrophysical Objects suite and the radiative transfer code Stellar Kinematics Including Radiative Transfer. We explore different subgrid post-processing recipes, using color distributions and their dependence on axis ratios of galaxies in the nearby Universe to tune and validate them. We find that subgrid post-processing recipes that mitigate limitations in the temporal and spatial resolution of the simulations are required for producing FUV to FIR photometry that statistically reproduce the colors of galaxies in the nearby Universe. With this paper, we release resolved photometry and spatially integrated spectra for our sample galaxies, each from a range of different viewing angles. Our simulations predict that there is a large variation in attenuation laws among galaxies, and that from any particular viewing angle that energy balance between dust attenuation and re-emission can be violated by up to a factor of 3. These features are likely to affect SED modeling accuracy.

Research on the Underlying Principles and Deep Learning Algorithms based on Image Style Conversion Techniques

Image style transformation techniques are an important research area in the field of computer vision, aiming to combine the content of one image with the artistic style of another image to create new images with the content of the input image and the style of the artistic style image. The development of this field has benefited from the rapid development of deep learning algorithms, especially the application of techniques such as Convolutional Neural Networks (CNN) and Generative Adversarial Networks (GAN). At the heart of image style transformation is how to capture both the content and artistic style of the input image. This is typically achieved by using different levels of activation of the convolutional neural network as content and style representations. The content representation captures the object and structural information in the image, while the style representation captures the texture and colour information of the image. To capture the style of an image, a Gram matrix is typically used to measure the correlation between features at a particular level. This matrix representation allows us to capture texture information and colour distribution, thus enabling style migration. The goal of the image style transformation model is to minimise both content loss and style loss to produce synthetic images. Content loss typically uses Euclidean distance to measure the difference between content representations, while style loss involves the distance between Gram matrices. GANs have become an important tool in image style transformation. The generator network is responsible for generating synthetic images and the discriminator network evaluates the realism of the generated images. Through adversarial training, the generator and discriminator networks constantly compete to improve the quality of the generated images. Finally, this study highlights the potential applications of image style conversion techniques in the fields of art creation, image editing and virtual reality, and proposes directions for future research on deep learning algorithms and underlying principles to further improve the efficiency and quality of image style conversion techniques. The continuous development of this field will bring new opportunities for image processing and creative applications.

Polymorphism of yak blood groups depending on coat color and collection

Yaks of the Republic of Tyva are characterised by high biological diversity, expressed by a wide range of body colours and polymorphism of blood groups. According to the distribution of colour and polymorphic blood systems in bulls and yak dams, the most characteristic colours are black, black-and-white, and brown; their frequency in the population is 0.403, 0.148 and 0.118, respectively. Animals of other colour patterns were encountered with a frequency ranging from 0.004 to 0.091. Differentiation of yak bulls and dams by the presence or absence of horns showed that in polled bulls and dams, the frequency of markers Q, L', F of the blood systems B, C, F-V is higher (from 0.250 to 0.846) than in horned ones (0.100–0.500). Blood markers R2, W, X1, U' systems C, and S were identified in polled bulls and dams with a frequency of 0.018–0.143, not found in horned bulls and dams. The frequency of blood antigens G2, T2, I2, E'2, Y', J2 of systems B, J is higher in horned bulls and dams (from 0.200 to 0.800) than in polled individuals (0.054–0.464), and S1 of blood system S was detected only in horned queens. Blood markers I1, Y2, R2, W, X1, and U' systems B, C, and S were undetected in horned bulls and dams. In polled and horned bulls, the frequency of genes E'2, E, X2, V, Z of blood systems B, C, F-V, and Z is higher; they are detected with a frequency of 0.333 to 1.000, regardless of colour. Polled bulls of grey colour do not have markers A2, G2, I1, O2, Y', Y2, Q', Y', X1 blood systems A, B, C, and brown colour - A2, O2, Q', F, U, U' blood systems A, B, F-V, S. Still, in polled bulls with black and brown colours, blood antigens I1, X1 were identified, which are absent in horned individuals of different colours. In yaks, statistically significant differences were noted in colour, polledness and antigens A2, G2, I1, I2, Q, T2, E'2, I', Y', R2, W, X1, X2, L', F, V, J2, H', U, U', Z.

Attention-based dual-path feature fusion network for automatic skin lesion segmentation

Automatic segmentation of skin lesions is a critical step in Computer Aided Diagnosis (CAD) of melanoma. However, due to the blurring of the lesion boundary, uneven color distribution, and low image contrast, resulting in poor segmentation result. Aiming at the problem of difficult segmentation of skin lesions, this paper proposes an Attention-based Dual-path Feature Fusion Network (ADFFNet) for automatic skin lesion segmentation. Firstly, in the spatial path, a Boundary Refinement (BR) module is designed for the output of low-level features to filter out irrelevant background information and retain more boundary details of the lesion area. Secondly, in the context path, a Multi-scale Feature Selection (MFS) module is constructed for high-level feature output to capture multi-scale context information and use the attention mechanism to filter out redundant semantic information. Finally, we design a Dual-path Feature Fusion (DFF) module, which uses high-level global attention information to guide the step-by-step fusion of high-level semantic features and low-level detail features, which is beneficial to restore image detail information and further improve the pixel-level segmentation accuracy of skin lesion. In the experiment, the ISIC 2018 and PH2 datasets are employed to evaluate the effectiveness of the proposed method. It achieves a performance of 0.890/ 0.925 and 0.933 /0.954 on the F1-score and SE index, respectively. Comparative analysis with state-of-the-art segmentation methods reveals that the ADFFNet algorithm exhibits superior segmentation performance.

A Deep Learning Image-to-Image Translation Approach for a More Accessible Estimator of the Healing Time of Burns.

An accurate and timely diagnosis of burn severity is critical to ensure a positive outcome. Laser Doppler imaging (LDI) has become a very useful tool for this task. It measures the perfusion of the burn and estimates its potential healing time. LDIs generate a 6-color palette image, with each color representing a healing time. This technique has very high costs associated. In resource-limited areas, such as low- and middle-income countries or remote locations like space, where access to specialized burn care is inadequate, more affordable and portable tools are required. This study proposes a novel image-to-image translation approach to estimate burn healing times, using a digital image to approximate the LDI. This approach consists of a U-net architecture with a VGG-based encoder and applies the concept of ordinal classification. Paired digital and LDI images of burns were collected. The performance was evaluated with 10-fold cross-validation, mean absolute error (MAE), and color distribution differences between the ground truth and the estimated LDI. Results showed a satisfactory performance in terms of low MAE ( 0.2370 ±0.0086). However, the unbalanced distribution of colors in the data affects this performance. This novel and unique approach serves as a basis for developing more accessible support tools in the burn care environment in resource-limited areas.

Breast Cancer Detection in Histopathology Images using ResNet101 Architecture

Cancer is a significant challenge in many fields, especially health and medicine. Breast cancer is among the most common and frequent cancers in women worldwide. Early detection of cancer is the main step for early treatment and increasing the chances of patient survival. As the convolutional neural network method has grown in popularity, breast cancer can be easily identified without the help of experts. Using BreaKHis histopathology data, this project will assess the efficacy of the CNN architecture ResNet101 for breast cancer image classification. The dataset is divided into two classes, namely 1146 malignant and 547 benign. The treatment of data preprocessing is considered. The implementation of data augmentation in the benign class to obtain data balance between the two classes and prevent overfitting. The BreaKHis dataset has noise and uneven color distribution. Approaches such as bilateral filtering, image enhancement, and color normalization were chosen to enhance image quality. Adding flatten, dense, and dropout layers to the ResNet101 architecture is applied to improve the model performance. Parameters were modified during the training stage to achieve optimal model performance. The Adam optimizer was used with a learning rate 0.0001 and a batch size of 32. Furthermore, the model was trained for 100 epochs. The accuracy, precision, recall, and f1-score results are 98.7%, 98.73%, 98.7%, and 98.7%, respectively. According to the results, the proposed ResNet101 model outperforms the standard technique as well as other architectures.

Effect of laser incidence angle on the femtosecond laser ablation characteristics of silicon carbide ceramics

Due to their excellent physical and chemical properties, silicon carbide ceramics have become representative high-performance materials in aerospace, energy power, biomedical, and other fields. However, as a typical hard and brittle material, silicon carbide ceramics possess incredibly high hardness and brittleness, which pose challenges to traditional mechanical processing methods. Laser processing, as a non-contact method, offers a wide processing range and high processing efficiency and has broad application prospects in the field of hard and brittle material processing. In this paper, laser ablations of silicon carbide ceramics with various combinations of process parameters were carried out by using a five-axis infrared femtosecond laser processing system. The changes in the ablation morphology with different incidence angles and laser power were discussed. The results demonstrate that the ablation threshold of silicon carbide ceramics initially increases, then decreases, and finally increases with the increase of laser incident angle. This trend may be attributed to the effect of the incidence angle on the laser absorption rate of silicon carbide ceramics. The ablation morphology of silicon carbide ceramics can be divided into three zones based on color and oxygen distribution: the ablation removal zone, ablation edge zone, and ablation affected zone. At the same laser incidence angle, the ablation range and depth gradually increase with the increase in laser power, but the growth rate gradually decreases. Furthermore, at the same laser power, as the incident angle increases, the laser energy density gradually decreases, and the long axis of the actual ablation area gradually lengthens while the short axis gradually shortens. This work could provides technical and theoretical support for laser surface processing at multiple incidence angles, which is conducive to multi-axis processing of complex curved parts.