Color Image Research Articles

Texture discrimination via Hilbert curve path based information quantifiers

The analysis of the spatial arrangement of colors and roughness/smoothness of figures is relevant due to its wide range of applications. This paper proposes a texture characterization method that extracts data from images using the Hilbert curve. Three information theory quantifiers are then computed: permutation entropy, permutation complexity, and Fisher information measure. The proposal exhibits some important properties: (i) it allows discrimination between figures according to varying degrees of correlations (as measured by the Hurst exponent), (ii) it is invariant to rotation and symmetry transformations, (iii) it is invariant to image scaling, (iv) it can be used for both black and white and color images. Validations have been performed not only using synthetic images but also using the well-known Brodatz image database.

Metasurface color filter arrays with a high efficiency and low color error

Conventional digital cameras combine absorbing color filter arrays with microlenses to achieve color imaging and improve efficiency. Such cameras require multi-step and multi-material fabrication processes. Several recent efforts have investigated metasurface-based color routing to combine focusing with filtering in a single functional layer with an improved efficiency. These approaches require high-refractive index materials and deep sub-micron fabrication to realize the metasurfaces. We present here an alternative, 2.5 dimensional metasurface that simultaneously provides both color filtering and focusing, but requires only a low-refractive index polymer and micron-scale patterning such that it is suitable for replication by molding. Unlike Bayer filters, this metasurface produces six independent spectra focused on nine monochrome pixels yielding both a high efficiency and low color error. These metasurfaces could be more photo-stable and thermally stable than dye-based filters and less expensive to produce than conventional arrays or metasurface color routers. Here, we characterize a metasurface-based focusing color filter array prototyped using two-photon lithography whose efficiencies are competitive with Bayer filters and whose color error is comparable to the limit of human perception.

Genome-wide association study on color-image-based convolutional neural networks.

Convolutional neural networks have excellent modeling abilities to complex large-scale datasets and have been applied to genomics. It requires converting genotype data to image format when employing convolutional neural networks to genome-wide association studies. Existing studies converting the data into grayscale images have shown promising. However, the grayscale image may cause the loss of information of the genotype data. In order to make full use of the information, we proposed a new method, color-image-based convolutional neural networks, by converting the data into color images. The experiments on simulation and real data show that our method outperforms the existing methods proposed by Yue and Chen for converting data into grayscale images, in which the model accuracy is improved by an average of 7.61%, and the ratio of disease risk genes is increased by an average of 18.91%. The new method has better robustness and generalized performance.

A Smart Camera With Integrated Deep Learning Processing for Disease Detection in Open Field Crops of Grape, Apple, and Carrot

ABSTRACTDowny mildew (Plasmopara), apple scab (Venturia inaequalis), and Alternaria leaf blight are endemic diseases that affect crops worldwide. The diseases can cause severe losses in grapes, apples and carrots when not detected and treated in an early stage. The European Union Horizon 2020 OPTIMA project aimed to improve disease detection in the open field with an automated detection system as part of an integrated pest management (IPM) system. In this research, we investigated the automated detection of downy mildew in grape, apple scab in apple and Alternaria leaf blight in carrot, using a deep convolutional neural network (CNN) on RGB color images. Detections from the CNN served as input to a Decision Support System (DSS), to precisely locate and quantify the disease, so that appropriate and timely application of plant protection products could be recommended. The focus of our study was on a smart camera implementation with integrated deep‐learning processing in real‐field conditions. The question was whether the deep learning model, when trained on images of disease symptoms recorded in conditioned circumstances, can also perform on images of disease symptoms recorded in field conditions. This type of evaluation is called open‐set evaluation, and so far it has received little attention in plant disease detection research. Therefore, the goal of our research was to evaluate the performance of a deep learning model in an open‐set evaluation scenario in commercial vineyards, orchards, and open fields. The model's performance in the open‐set scenario was compared to its performance in the closed‐set scenario, which involved evaluating the trained model on images similar to those used for model training. Our results showed that the model's performance in the closed‐set scenario with F1 scores of 66.3% (downy mildew), 45.1% (apple scab), and 42.1% (Alternaria) was notably better than in the open‐set scenario, with F1 scores of 34.8% (downy mildew), 5.5% (apple scab) and 4.2% (Alternaria). Uniform Manifold Approximation and Projection (UMAP) analysis proved the significant difference between the open‐set and closed‐set data sets. Our result should encourage other researchers to carry out similar open‐set evaluations to get realistic impressions of their model's performance under field conditions. A subset of our image data set has been made publicly available at https://doi.org/10.5281/zenodo.6778647.

Application of Computer Vision in Automatic Generation of Product Appearance Design

One of the most important aspects of industrial manufacturing is the assessment of product appearance and quality, which frequently occurs manually. A product’s appearance can convey practical qualities, aesthetic value, and a sense of quality to consumers. This study explores the use of computer vision in the automatic generation of product appearance designs. In this study, a dolphin swarm-driven weighted random forest (DS-WRF) design approach is introduced to assist designers in improving the efficiency of product appearance design with automatic generation. In this study, we collected product image data. The data were preprocessed to resize the images. The two primary components of this method are sketched inversion and design scheme creation. The produced sketches are entered into the sketch inversion model combined with the matching color images for training. Sketches by hand are transformed into color design patterns using this approach. The experimental result demonstrates that the proposed method involves the automatic generation of product appearance designs with high quality.

Identification of Papaya Ripeness Using the Support Vector Machine Algorithm

Papaya is a tropical fruit that is commonly consumed and found in Indonesia. The ripeness level of papaya is typically assessed based on its colour. However, farmers and consumers often make mistakes identifying the fruit's ripeness. This research aims to design an application capable of determining the ripeness level of papaya based on colour images using Red, Green, Blue (RGB) and Hue, Saturation, Value (HSV) features and applying the Support Vector Machine (SVM) algorithm for ripeness classification. The dataset consists of images of California papayas, with 150 samples. The outcome of this study is a digital image application that can classify papaya ripeness into three categories: raw, half-ripe, and fully ripe. The evaluation used 80% of the data for training and 20% for testing. The results show an accuracy of 80%. With this relatively high level of accuracy, it can be concluded that the SVM algorithm is reliable for classifying papaya ripeness levels of Papayas.

Reconfigurable and nonvolatile ferroelectric bulk photovoltaics based on 3R-WS2 for machine vision

Hardware implementation of reconfigurable and nonvolatile photoresponsivity is essential for advancing in-sensor computing for machine vision applications. However, existing reconfigurable photoresponsivity essentially depends on the photovoltaic effect of p-n junctions, which photoelectric efficiency is constrained by Shockley-Queisser limit and hinders the achievement of high-performance nonvolatile photoresponsivity. Here, we employ bulk photovoltaic effect of rhombohedral (3R) stacked/interlayer sliding tungsten disulfide (WS2) to surpass this limit and realize highly reconfigurable, nonvolatile photoresponsivity with a retinomorphic photovoltaic device. The device is composed of graphene/3R-WS2/graphene all van der Waals layered structure, demonstrating a wide range of nonvolatile reconfigurable photoresponsivity from positive to negative ( ± 0.92 A W−1) modulated by the polarization of 3R-WS2. Further, we integrate this system with a convolutional neural network to achieve high-accuracy (100%) color image recognition at σ = 0.3 noise level within six epochs. Our findings highlight the transformative potential of bulk photovoltaic effect-based devices for efficient machine vision systems.

Data Shuffling and Data Blocks Shuffling Method for Digital Data Signals Protection

In this paper, a new method for digital data cryptography is proposed to increase the speed of data cryptography, simplify the processes of data encryption - decryption, and to strengthen the degree of data protection. The proposed data shuffling and data blocks shuffling (DS_DBS) method is able to process messages, gray images, color images and digital speech file using the same encryption and decryption functions; changing the digital data type will not require any changes in the aforesaid functions. The proposed DS_DBS method allows data blocking (block size will be variable), and it is determined by the private key. Data encryption is applied by simple data shuffling and data blocks shuffling, while data decryption is applied by shuffling back the data blocks and shuffling back the data. These shuffling operations replace the complex logical and arithmetic operations used in other existing methods of data cryptography. The shuffling and shuffling back operations are implemented based on two secret indices keys, obtained by running two chaotic logistic map models. The proposed method utilizes a long private key with a variable length that depends on the selected number of crypto phases. It can be implemented in one or more phases; each phase is a function call to execute the encryption-decryption functions with the associated inputs. Using more than a phase, increases the security level by using longer private key, and each phase will be independent. The encrypted-decrypted result of each phase can be taken as a final result. Implementing the proposed DS_DBS method - using various data types – and examining its speed reveal not only the enhanced speed of data cryptography, but also the better quality and sensitivity of the proposed method compared to those of existing state of the art cryptography methods.

Color image segmentation method based on extended active membrane system

Color image segmentation method based on extended active membrane system

Significance of Endoscopic Redness of Duodenum in Health Checkup

ABSTRACTBackground and AimIt is important for endoscopist to diagnose the lesion redness. In this study, we focused on the redness of duodenal bulb. We objectively analyzed the changes in redness of the duodenal bulb using linked color imaging (LCI) with chromatic indicators.MethodsSeven endoscopists observed the duodenal bulb with white light imaging (WLI) and LCI, and evaluated them by visual analogue scale (VAS) for the degree of redness. The difference in VAS between WLI and LCI was defined as ΔVAS. All images were quantified by the Comission Internationale de l'Eclariage‐L*a*b* color space. Values related to color differences (ΔE*, ΔL*, Δa*, and Δb*) were calculated from the two images of WLI and LCI. Multiple regression analysis was performed for the factors with the health checkup correlated with ΔVAS and the correlation between ΔVAS and ΔE*, ΔL*, Δa*, and Δb* was also examined.ResultsThe analysis prospectively included 1144 examinees. In multiple regression analysis, it revealed that sex (β = 0.5847, p < 0.0001) and metabolic syndrome (β = 0.4138, p = 0.0012) were the factors independently influenced ΔVAS. And only Δa*, a chromatic index for changes in the degree of redness, showed a statistically and considerably positive correlation with ΔVAS (r = 0.4529, p < 0.0001).ConclusionTo evaluate the difference in the degree of redness between WLI and LCI of duodenal bulb in esophagogastroduodenoscopy may help in early detection of metabolic syndrome, which rarely has symptoms.

Simple and rapid paper-based colorimetric spot test for promethazine screening in recreational drug beverages and pharmaceuticals

This study introduces a paper-based colorimetric spot test for the rapid and cost-effective detection of promethazine, featuring enhanced stability and user-friendliness. The developed device employs a potassium persulfate reagent encapsulated in a xerogel matrix, facilitating a simple chemical oxidation reaction that produces a pink color upon promethazine detection. Color quantification is efficiently achieved using a digital camera, with image analysis performed through ImageJ software to convert color image into a blue intensity scale. Under optimized conditions, the method demonstrates a linear calibration curve within the range of 100–600 mg/L and a detection limit of 20 mg/L. The paper-based colorimetric spot test exhibited good intra-day and inter-day accuracy of the recoveries of 98–102 % and 99–101 %, respectively. Good precision of measurement was also achieved (5.6 % RSD). The xerogel matrix ensures the long-term storage stability of the reagent for three months when stored at both 25 °C and 4 °C, making the device suitable for field applications. When applied to detection of promethazine in recreational drug beverage samples and pharmaceutical formulations, the performance of the proposed method showed no significant statistical differences compared to gas chromatography–flame ionization detection (GC–FID). This low-cost, portable, and disposable tool is suited for onsite screening of promethazine, representing a significant advancement in rapid drug detection methodologies.

SAR Image Colorization with a Deep Learning Model for All-Inclusion Understanding (H)

The procedure for enhancing halftone photographs with color to create eye-catching pigmented pictures is known as image colorization. This involves the application of colors in line with the intended purpose of the image. However, traditional methods of colorization available today often include the need for end user’s intervention in the form providing color points and doodle drawings, as well as use of color images for reference in which case colors are transferred to the target image or using multiple color images to predict what the colored result would look like; however, the majority of colorizes developed tend to produce colorized images that are not realistic in most instances due to unskilled users who attempt to apply the devices, wrong color transfer or limited color image library. In this paper, to overcome these weaknesses, a new method that consists of two modules is proposed to colorize certain regions of an image by fusing semantic segmentation and seamless region filling. In the first module, input image foreground and background regions masks and categories are extracted each extracted region's reference image is taken from a per-sorted color image database using a Mask R-CNN model. Using colorization approaches based on the VGG and U-Net models, respectively, the second module entails coloring the image's backdrop and other areas. Afterward, the final, fully colored image is created by combining the photographs using the poission editing process. The tests show that our method enables not only the proper reference image selection based on the semantic information of various image components, but also incorporation of the colored results to create a believable colored image. Proposing to use different CNNbased models in a combination where each is used according to its own strength, In contrast to other methods now in use, our scheme gives a deeper creative visual impression while avoiding the disadvantage of single-step techniques' failure frequency. Keywords: Mask R-CNN, Input Photograph, Image Sections, Coloured Photographs, Conceptual Data, Visuals, Black and White Images, Semantic Pix elation, Reference Picture, Base Image, Images Stockpile, Outlying Areas, Image Library, Segregation Methods, Doodle, Conventional Neural Network Based Designs, U-Net Architecture, One-shot Approach, Strategy Present in This Manuscript, Zonal Regions, Principal Zones, Large Image, Color space, Competitive Half toning Algorithm, Zonal Context, Focus Area, Artistic Incorporation, Partially Automatic Process, Cutting Out Section of the Picture, Instance Level Segmentation

New S-Box Generation Based on Hybrid Two-Dimensional Chaotic Map for Color Image Encryption

New S-Box Generation Based on Hybrid Two-Dimensional Chaotic Map for Color Image Encryption

Postfire Estimation of Heating Temperatures Experienced by Fire Retardant Coatings Using Smartphone Videos and Machine Learning

ABSTRACTAccurate estimation of heating temperatures experienced by fire retardant coatings (FRCs) is crucial in identifying the ignition source during fire investigations. While traditional methods, such as spectroscopy, effectively capture the compositional changes in FRC at various heating temperatures, they are typically bulky, costly, and unsuitable for rapid field analysis. This study proposes the use of smartphone and machine learning to predict the heating temperatures of FRC. A smartphone is employed to capture short videos of FRC samples illuminated by its color‐changing screen. Video frames are then decomposed into color images and converted into spectral data for further processing. Linear and nonlinear regression models are applied to identify key variables and enhance predictive accuracy. The performance of smartphone‐based temperature estimation is compared to that of hyperspectral imaging and laser‐induced breakdown spectroscopy. In the test phase, the coefficient of determination for smartphone‐based estimation ranges from 0.946 to 0.962, often surpassing that of benchmark methods. These results indicate that smartphones can provide a low‐cost, effective means for estimating heating temperatures of FRC in fire investigations.

Metasurface‐Enabled Cascaded Multispectral Polarization Structures Along the Longitudinal Direction

AbstractGeneration and manipulation of polarization structures have attracted significant interest due to their unusual optical features and extensive applications. Multispectral information can further increase the information capacity of polarization structures. However, generating multiple cascaded multispectral polarization numbers (letters) at multiple planes arranged along light propagation has not been reported. A geometric metasurface is used to experimentally realize eight cascaded multispectral polarization numbers (letters) at predesigned observation planes along the longitudinal direction. The efficacy of this approach is demonstrated by encoding a single polarization number (letter) with two wavelengths and cascading two polarization numbers with orthogonal polarization rotation angles. Different two‐color polarization numbers (letters) are generated by controlling the incident wavelength, while their polarization distributions are modulated by controlling the incident linear polarization. This approach integrates two‐dimentional (2D) polarization generation, multiple colors, longitudinal control, and cascading design, providing extra degrees of freedom for customized polarization design and enabling small‐volume polarization systems for color display, image steganography, and optical encryption.

Anisotropic Diffusion in Riemannian Colour Geometry

Anisotropic diffusion has long been an important tool in image processing. More recently, it has also found its way to colour imaging. Until now, mainly Euclidean colour spaces have been considered in this context, but recent years have seen a renewed interest in and importance of non-Euclidean colour geometry. The main contribution of this paper is the derivation of the equations for anisotropic diffusion in Riemannian colour geometry. It is demonstrated that it contains several well-known solutions such as Perona–Malik diffusion and Tschumperlé–Deriche diffusion as special cases. Furthermore, it is shown how it is non-trivially connected to Sochen’s general framework for low-level vision. The main significance of the method is that it decouples the coordinates used for solving the diffusion equation from the ones that define the metric of the colour manifold, and thus directs the magnitude and direction of the diffusion through the diffusion tensor. It also enables the use of non-Euclidean colour manifolds and metrics for applications such as denoising, inpainting, and demosaicing, based on anisotropic diffusion.

An Overview of Quantum Circuit Design Focusing on Compression and Representation

Quantum image computing has attracted attention due to its vast storage capacity and faster image data processing, leveraging unique properties such as parallelism, superposition, and entanglement, surpassing classical computers. Although classical computing power has grown substantially over the last decade, its rate of improvement has slowed, struggling to meet the demands of massive datasets. Several approaches have emerged for encoding and compressing classical images on quantum processors. However, a significant limitation is the complexity of preparing the quantum state, which translates pixel coordinates into corresponding quantum circuits. Current approaches for representing large-scale images require higher quantum resources, such as qubits and connection gates, presenting significant hurdles. This article aims to overview the pixel intensity and state preparation circuits requiring fewer quantum resources and explore effective compression techniques for medium and high-resolution images. It also conducts a comprehensive study of quantum image representation and compression techniques, categorizing methods by grayscale and color image types and evaluating their strengths and weaknesses. Moreover, the efficacy of each model’s compression can guide future research toward efficient circuit designs for medium- to high-resolution images. Furthermore, it is a valuable reference for advancing quantum image processing research by providing a systematic framework for evaluating quantum image compression and representation algorithms.

Residual Vision Transformer and Adaptive Fusion Autoencoders for Monocular Depth Estimation.

Precision depth estimation plays a key role in many applications, including 3D scene reconstruction, virtual reality, autonomous driving and human-computer interaction. Through recent advancements in deep learning technologies, monocular depth estimation, with its simplicity, has surpassed the traditional stereo camera systems, bringing new possibilities in 3D sensing. In this paper, by using a single camera, we propose an end-to-end supervised monocular depth estimation autoencoder, which contains an encoder with a structure with a mixed convolution neural network and vision transformers and an effective adaptive fusion decoder to obtain high-precision depth maps. In the encoder, we construct a multi-scale feature extractor by mixing residual configurations of vision transformers to enhance both local and global information. In the adaptive fusion decoder, we introduce adaptive fusion modules to effectively merge the features of the encoder and the decoder together. Lastly, the model is trained using a loss function that aligns with human perception to enable it to focus on the depth values of foreground objects. The experimental results demonstrate the effective prediction of the depth map from a single-view color image by the proposed autoencoder, which increases the first accuracy rate about 28% and reduces the root mean square error about 27% compared to an existing method in the NYU dataset.

Stable and Lead‐Free Perovskite Hemispherical Photodetector for Vivid Fourier Imaging

AbstractThe filterless single‐pixel imaging technology is anticipated to hold tremendous competitiveness in diverse imaging applications. Nevertheless, achieving single‐pixel color imaging without a filter remains a formidable challenge. Here a lead‐free perovskite hemispherical photodetector is reported for filterless single‐pixel color imaging. Through passivating the active vacancy and antisite defects with functional chemical groups, the trap density of the perovskite film can be dramatically reduced to half of the pristine one, which results in a stable device with 0.69 nA standard deviation (SD) of the current fluctuations for over 10 000 s. The exceptional stability and remarkable low light response capability of passivated lead‐free perovskite devices ensure reliable and repeated distinguishment to high‐frequency color information of Fourier patterns. High‐resolution color imaging with 256 × 256 pixels can be achieved by integrating the Bayer template with Fourier patterns and designing a corresponding color restoration algorithm. Notably, its imaging quality surpasses commercial silicon photomultiplier tubes (SiPMs) under low light intensities or high‐temperature conditions as demonstrated by the larger signal‐to‐noise ratio (SNR). This work presents a novel approach for applying perovskite photodetectors in single‐pixel color imaging, propelling the advancement of filterless color imaging technology.

Image use in aphasia rehabilitation: Practices of speech-language pathologists in South Africa and international contexts.

The study aimed to explore speech-language pathologists' practices regarding image selection and use in the treatment of people living with aphasia. Images are widely utilised with people living with aphasia, however, supporting their communication optimally remains challenging. Ninety-two speech-language pathologists from South Africa, Australia, USA, and the UK completed a custom-designed online survey. Descriptive and inferential statistics, with structured-tabular thematic analysis were used. Sixty-two (82.7%) of the 75 participants who conduct assessments with people living with aphasia mostly use images. Sixty-seven (72.8%) of the total sample include images during treatment. South African respondents reported using black-and-white line drawings (28.1%) less frequently than speech-language pathologists from other countries (9.8%). Coloured images were preferred overall. South African participants significantly reported culture (62.5%) and religious orientation as "always important" (63.3%) compared to their counterparts (34.1%, 12.2%). Respondents use images to support learning and language; and less frequently in discourse with people living with aphasia. The more experienced a respondent was, the more knowledgeable and confident they felt incorporating images. Themes of client considerations and clinical settings for images emerged. Speech-language pathologists' perceived barriers were time, larger caseloads, people living with aphasia's variable responses to images, and lack of resources. Speech-language pathologists in this sample have a knowledge base of image types and factors that may contribute to image interpretation in people living with aphasia. However, further research globally is needed. A structured approach to image selection is lacking in both low and middle-income, and high-income countries. Considering people living with aphasia's skills, preferences, and communication partner involvement may optimise image use. Development of a framework for systematically organising image features for people living with aphasia is warranted.