Pseudo-color Images Research Articles

On properties of visual quality metrics in remote sensing applications

Visual quality is important for remote sensing data presented as grayscale, color or pseudo-color images. Although several visual quality metrics (VQMs) have been used to characterize such data, only a limited analysis of their applicability in remote sensing applications has been done so far. In this paper, we study correlation factors for a wide set of VQMs for color images with distortion types typical for remote sensing. It is demonstrated that there are many metrics that have very high Spearman rank order correlation, e.g. PSNR-based and SSIM-based metrics. Meanwhile, there are also metrics that are practically uncorrelated with others. A detailed analysis of VQMs that have the largest SROCC values and belong to different groups is presented in this paper.

Vehicle Trace Detection in Two-Pass SAR Coherent Change Detection Images With Spatial Feature Enhanced Unet and Adaptive Augmentation

As a typical application of remote sensing technology, change detection can find the ground information changes by acquiring images of the same region at different times. The change detection using the synthetic aperture radar (SAR) with the advantages of all day and all-weather usually monitors the significant surface change, like flood disasters and earthquake deformation. However, when it comes to detecting subtle changes like vehicle traces, the traditional methods ignoring the phase coherence between image pairs cannot intensify these faint changes in the difference image. The SAR coherent change detection (CCD) based on repeat-pass repeat-geometry complex images utilizing both the intensity and phase fraction could exhibit the subtle vehicle trace in the difference image. However, the complicated background and decorrelation factors significantly affect the quality of difference images, further causing great trouble for automatic trace detection. This paper proposes the spatial feature enhanced Unet and adaptive data augmentation to realize vehicle trace detection. More specifically, the pseudo-color image is first synthesized based on a two-stage coherence estimation method. Then considering the long-continuity and parallel distribution of vehicle trace samples, the enhanced Unet is constructed by fusing spatial convolutional neural network and spatial attention mechanism. After that, the adaptation data augmentation strategy is presented by introducing manual registration errors and multiple estimation windows. Finally, the experimental results on the Sandia CCD data and our measured data demonstrate the effectiveness of the proposed method.

PolSAR-SSN: An End-to-End Superpixel Sampling Network for PolSAR Image Classification

Polarimetric synthetic aperture radar (PolSAR) image classification is one of the fundamental research areas in remote sensing. Superpixels can provide boundary constraint information and are widely used in PolSAR image interpretation. However, traditional machine learning superpixel algorithms have many limitations for PolSAR image interpretation. Pseudo-color images are usually used as the superpixel algorithm inputs, and the loss of polarimetric information will decrease the performance. In addition, the superpixel algorithms are difficult to incorporate into state-of-the-art deep learning models and cannot be trained in an end-to-end manner. In this letter, a trainable end-to-end deep superpixel network is proposed for PolSAR image classification. The inputs of the proposed method can be any low/middle-level polarimetric features of a PolSAR image and the rich polarimetric feature representation can be learned. The produced superpixels of the proposed method are more concentrated near the land cover boundaries and can significantly improve the performance of PolSAR image classification. Experimental results show that the overall accuracies of the proposed method are approximately 2.57% and 1.44% higher than traditional superpixel algorithms on two PolSAR datasets and surpass some well-known deep learning methods.

Multistage Spatio-Temporal Networks for Robust Sketch Recognition.

Sketch recognition relies on two types of information, namely, spatial contexts like the local structures in images and temporal contexts like the orders of strokes. Existing methods usually adopt convolutional neural networks (CNNs) to model spatial contexts, and recurrent neural networks (RNNs) for temporal contexts. However, most of them combine spatial and temporal features with late fusion or single-stage transformation, which is prone to losing the informative details in sketches. To tackle this problem, we propose a novel framework that aims at the multi-stage interactions and refinements of spatial and temporal features. Specifically, given a sketch represented by a stroke array, we first generate a temporal-enriched image (TEI), which is a pseudo-color image retaining the temporal order of strokes, to overcome the difficulty of CNNs in leveraging temporal information. We then construct a dual-branch network, in which a CNN branch and a RNN branch are adopted to process the stroke array and the TEI respectively. In the early stages of our network, considering the limited ability of RNNs in capturing spatial structures, we utilize multiple enhancement modules to enhance the stroke features with the TEI features. While in the last stage of our network, we propose a spatio-temporal enhancement module that refines stroke features and TEI features in a joint feature space. Furthermore, a bidirectional temporal-compatible unit that adaptively merges features in opposite temporal orders, is proposed to help RNNs tackle abrupt strokes. Comprehensive experimental results on QuickDraw and TU-Berlin demonstrate that the proposed method is a robust and efficient solution for sketch recognition.

МЕТОДИ ПІДВИЩЕННЯ ІНФОРМАТИВНОСТІ ТА ЗМЕНШЕННЯ ОБ'ЄМУ ГРАФІЧНИХ ДАНИХ НА ОСНОВІ АНАЛІЗУ ЇХ КОЛІРНОГО ПРОСТОРУ

The constant development of digital technology has led to a sharp increase in the number and volume of media files, including digital images, which make up a significant part of computer network traffic, which reduces the speed of their transmission. The research conducted in this work is based on the provisions and methods of digital image processing, the laws of visual perception, the basics of probability theory and mathematical modeling. The results of theoretical research were verified by simulation. The paper proposes a technology that, through the analysis of the color space of the image and taking into account the laws of visual perception, makes it possible to significantly reduce the size of the image file. This technology is used to solve a number of problems, in particular, the visualization of large files and increase the informativeness of images with complex semantic content. It is established that the reduction of the image file size is achieved through the optimization of the palette and leads to a slight deterioration in the visual quality of image perception. To reduce the visibility of error and create a visual sense of the presence of more different colors in the image than is actually the case, it is proposed to use diffuse pseudo-mixing of colors, which is to model some colors with others. Along with the task of reducing the volume of graphic files based on the optimization of the palette, a similar methodology was developed to increase the informativeness of images through the use of pseudo-colors. By modifying the function of converting the coordinates of color space into color components, a modified approach to the formation of pseudo-color images is proposed, which increases the informativeness of halftone digital images in their visual analysis.

A Dual-Polarimetric SAR Ship Detection Dataset and a Memory-Augmented Autoencoder-Based Detection Method.

With the development of imaging and space-borne satellite technology, a growing number of multipolarized SAR imageries have been implemented for object detection. However, most of the existing public SAR ship datasets are grayscale images under single polarization mode. To make full use of the polarization characteristics of multipolarized SAR, a dual-polarimetric SAR dataset specifically used for ship detection is presented in this paper (DSSDD). For construction, 50 dual-polarimetric Sentinel-1 SAR images were cropped into 1236 image slices with the size of 256 × 256 pixels. The variances and covariance of both VV and VH polarization were fused into R,G,B channels of the pseudo-color image. Each ship was labeled with both a rotatable bounding box (RBox) and a horizontal bounding box (BBox). Apart from 8-bit pseudo-color images, DSSDD also provides 16-bit complex data for readers. Two prevalent object detectors R3Det and Yolo-v4 were implemented on DSSDD to establish the baselines of the detectors with the RBox and BBox respectively. Furthermore, we proposed a weakly supervised ship detection method based on anomaly detection via advanced memory-augmented autoencoder (MemAE), which can significantly remove false alarms generated by the two-parameter CFAR algorithm applied upon our dual-polarimetric dataset. The proposed advanced MemAE method has the advantages of a lower annotation workload, high efficiency, good performance even compared with supervised methods, making it a promising direction for ship detection in dual-polarimetric SAR images. The dataset is available on github.

Early Assessment of Response to Radiofrequency Ablation With CT Perfusion Imaging in Rabbit VX2 Liver Tumor Model.

ObjectivesTo discriminate viable tumors from benign periablational enhancement (BPE) in early stage after radiofrequency ablation (RFA) is a major confounding problem. The goal of this study is to evaluate quantitative assessment and diagnostic value of CT perfusion between viable tumors and BPE after RFA in the rabbit liver VX2 tumor model, with pathological results as the standard.MethodsTwenty-eight VX2 liver tumors were treated with RFA, on days 1, 3, 7, and 14, seven rabbits were randomly chosen for CT perfusion and performed pathology examinations immediately. The perfusion parameters along with the profile of time-density curves (TDCs) and pseudo-color images of the parameters were observed in both BPE and viable tumors, then compared with the pathology results. The perfusion parameters included blood flow (BF), blood volume (BV), time to peak (TTP), permeability (P), arterial liver perfusion (ALP), portal venous perfusion (PVP) and hepatic perfusion index (HPI).ResultsA total of 26/28 rabbits successfully underwent CT perfusion, while 6/26 lesions were confirmed to be viable tumors. The TDCs of BPE were mainly speed-up platform curves (15/26), while the viable tumors showed mainly speed-up speed-down (3/6) and speed-up platform (2/6) curves. The PVP values were significantly higher, and the HPI values were significantly lower for BPE at all time points than viable tumors (P < 0.05). Both of PVP value and HPI value have high efficiency for the differential diagnosis of the viable tumors and BPE at each time point. These characteristics of CT perfusion parameters were consistent with pathological changes.ConclusionsThe TDCs, PVP and HPI have the potential to indicate BPE and viable tumors effectively early after RFA treatment, the results were highly consistent with pathology. CT perfusion has advantages with great efficacy in monitoring the therapeutic effect early after RFA treatment.

Statistical analysis of infrared thermogram for CNN-based electrical equipment identification methods

ABSTRACT It is essential to develop infrared (IR) thermogram identification technologies to establish automatic diagnosis systems in power substations. The convolutional neural network (CNN) based methods show the highest accuracy in this field. The IR thermograms of electrical equipment are very different from general digital images, which means the present methods need further improvements. For data-driven CNN methods, it is necessary to study the characteristics of the IR data. This paper collected 11817 thermograms from substations and structured the dataset according to equipment types. The statistical features of mean, variance, skewness, kurtosis and contrast are analyzed and compared with other five image datasets. Several tricks are revealed from the analysis and tested on CNN models. Firstly, greycaling the Iron pseudo-color images extracts the temperature information and makes it possible to design models with fewer channels. The test shows it could reduce over 35% computational costs. Secondly, the sparse information of color and edges of thermograms makes it necessary to keep the original aspect ratio. The image preprocessing method of cropping shows better performance than padding and rescaling. Thirdly, the 0–1 normalization can boost the training process for about 100 epochs, which is related to the particular background of thermograms.

Post-mortem evaluation of the pathological degree of myocardial infarction by Fourier transform infrared microspectroscopy

In clinical and forensic investigations, accurate post-mortem diagnosis of the pathological degree of myocardial infarction (MI) is critical. However, because of the observer variability, the diagnosis cannot be made objectively. Many studies have shown that Fourier transform infrared (FTIR) microspectroscopy is non-invasive, observer-independent, and label-free when analyzing biological tissues. In this study, we used FTIR microspectroscopy in combination with intelligent algorithms to identify the pathological phases in human infarcted cardiac tissues, including ischemia, necrotic, granulation, and fibrotic stages. First, a comparison of infrared spectra corresponding to infarcted tissue pathological categories revealed various spectral properties. The results of unsupervised principal component analysis (PCA) revealed a clear distinction between these four pathological stages and the normal stage. Then, to identify these five stages, an automatic artificial neural network (ANN) classifier was effectively created. Finally, two-dimensional pseudo-color images of two infarcted cardiac tissue sections visualized via the ANN classifier showed great agreement with their histological images. These findings demonstrate that FTIR microspectroscopy has the potential for the post-mortem evaluation of the pathological degree of MI.

Seam welds quality control improvement using method of pseudo-color coding images

The method of seam welds quality control using X-ray is considered. The X-ray methods of control are based on the capability of gamma radiation to penetrate through a metal including welded areas. Regions having defects - pores, faulty welds, cracks, scale inclusions - look darker on images. Appearance, linear dimensions and depths of the defects usually are determined by a visual examination comparing the X-ray image with standard defects images. It is known that a human eye can distinguish not more than 12-15 shades on a black and white image but more than a hundred on a colored image. The paper considers possibilities of the developed method by the author and based on the optical mixing of two or three complementary colors - red, blue and green. The method can use only one pair of the colors at a time, i.e. it is possible to have three various pairs for a pseudo-color image. The obtained pseudo-color image has the same informational capacity as the original black and white image. But the greater fraction of the saved information becomes available for visual examination of the X-ray image. In the end the efficiency of the seam weld quality control increases.

Painting and calligraphy identification method based on hyperspectral imaging and convolution neural network

It is of great social value and cultural and technological innovation demonstration value to carry out the research on the scientific identification method of painting and calligraphy works of art, and it is of great effect to the trade, collection, and protection of painting and calligraphy works of art. Spectral imaging, as an information acquisition method of attribute and visual synchronous perception, can be used for painting and calligraphy identification. In particular, through hyperspectral imaging and data analyses, we can identify the pigment ink used in painting, judge the printing characteristics, and find the painting information invisible to human eyes, to comprehensively judge the authenticity and abnormality of painting. However, due to its lack of matching painting and calligraphy identification model and algorithm, as well as special painting and calligraphy atlas database support, it is difficult to be competent for rapid and accurate identification in practice. Because of the above problems, in this paper, it is simulated the expert identification process for artificial intelligence analysis and modeling, adopts the idea of combining hyperspectral imaging and Atlas intelligent learning, proposes a method of atlas feature extraction for calligraphy and painting identification, and designs and studies convolution neural network(CNN) based on atlas feature, traditional image feature, and the mixed feature of atlas and image, to judge the authenticity of calligraphy and painting, the author and so on. The actual test results show that the convolution neural network based on the atlas features is the best, the author classification accuracy and authenticity identification rate in the test sample set are more than 96.5%, and it can also be seen that in the pseudo color image data, adding multivariate spectral features can greatly improve the accuracy significantly.

Development of a multi-excitation fluorescence (MEF) imaging method to improve the information content of benthic coral reef surveys

Benthic surveys are a key component of monitoring and conservation efforts for coral reefs worldwide. While traditional image-based surveys rely on manual annotation of photographs to characterise benthic composition, automatic image annotation based on computer vision is becoming increasingly common. However, accurate classification of some benthic groups from reflectance images presents a challenge to local ecologists and computers alike. Most coral reef organisms produce one or a combination of fluorescent pigments, such as Green Fluorescent Protein (GFP)-like proteins found in corals, chlorophyll-a found in all photosynthetic organisms, and phycobiliproteins found in red macroalgae, crustose coralline algae (CCA) and cyanobacteria. Building on the potential of these pigments as a target for automatic image annotation, we developed a novel imaging method based on off-the-shelf components to improve classification of coral and other biotic substrates using a multi-excitation fluorescence (MEF) imaging system. We used RGB cameras to image the fluorescence emission of coral and algal pigments stimulated by narrow-waveband blue and green light, and then combined the information into three-channel pseudocolour images. Using a set of a priori rules defined by the relative pixel intensity produced in different channels, the method achieved successful classification of organisms into three categories based on the dominant fluorescent pigment expressed, facilitating discrimination of traditionally problematic groups. This work provides a conceptual foundation for future technological developments that will improve the cost, accuracy and speed of coral reef surveys.

Color night vision ghost imaging based on a wavelet transform.

Night vision imaging is a technology that converts objects not visible to the human eye into visible images for night scenes and other low-light environments. However, conventional night vision imaging can directly produce only grayscale images. Here, we report a novel, to the best of our knowledge, color night vision imaging method based on a ghost imaging framework and optimized coincidence measurement based on wavelet transformation. An interesting phenomenon is that color night vision image can be directly produced by this new method. To our knowledge, this is the first direct color night vision imaging method without any conventional pseudocolor image fusion techniques. The experimental results show that this method can restore color very well for some objects. Moreover, the color of the night vision image is more natural and friendly to the human eye than that of conventional color night vision images. Due to the advantages of wavelet transforms, this method has high reconstruction ability for distorted signals.

High-throughput fast full-color digital pathology based on Fourier ptychographic microscopy via color transfer

Full-color imaging is significant in digital pathology. Compared with a grayscale image or a pseudo-color image that only contains the contrast information, it can identify and detect the target object better with color texture information. Fourier ptychographic microscopy (FPM) is a high-throughput computational imaging technique that breaks the tradeoff between high resolution (HR) and large field-of-view (FOV), which eliminates the artifacts of scanning and stitching in digital pathology and improves its imaging efficiency. However, the conventional full-color digital pathology based on FPM is still time-consuming due to the repeated experiments with tri-wavelengths. A color transfer FPM approach, termed CFPM was reported. The color texture information of a low resolution (LR) full-color pathologic image is directly transferred to the HR grayscale FPM image captured by only a single wavelength. The color space of FPM based on the standard CIE-XYZ color model and display based on the standard RGB (sRGB) color space were established. Different FPM colorization schemes were analyzed and compared with thirty different biological samples. The average root-mean-square error (RMSE) of the conventional method and CFPM compared with the ground truth is 5.3% and 5.7%, respectively. Therefore, the acquisition time is significantly reduced by 2/3 with the sacrifice of precision of only 0.4%. And CFPM method is also compatible with advanced fast FPM approaches to reduce computation time further.

A High-Resolution Microfluidic LAPS Imaging System for 3D Cell Metabolism Monitoring

A High-Resolution Microfluidic LAPS Imaging System for 3D Cell Metabolism Monitoring

Detection of early decay on citrus using LW-NIR hyperspectral reflectance imaging coupled with two-band ratio and improved watershed segmentation algorithm

Decay is a serious problem in citrus storage and transportation. However, the automatic detection of decayed citrus remains a problem. In this study, the long wavelength near-infrared (LW-NIR) hyperspectra reflectance images (1000–1850 nm) of oranges were obtained, and an effective method to detect decayed citrus was proposed. Three effective wavelength selection algorithms and two classification algorithms were used to build decay detection models in pixel-level, as well as the two-band ratio images, pseudo-color image enhancement and improved watershed segmentation were used to build decay detection models in image-level. The image-level detection method proposed in this study obtained a total success rate of 92% for all fruit, indicating its potential to detect decayed oranges online. Moreover, the LW-NIR hyperspectral reflectance imaging is verified as a useful method to detect surface defects of fruits.

Hyperpigmented spots at fundus examination: a new ocular sign in Neurofibromatosis Type I

BackgroundNeurofibromatosis Type I (NF1), also termed von Recklinghausen disease, is a rare genetic disorder that is transmitted by autosomal dominant inheritance, with complete penetrance and variable expressivity. It is caused by mutation in the NF1 gene on chromosome 17 encoding for neurofibromin, a protein with oncosuppressive activity, and it is 50% sporadic or inherited. The disease is characterized by a broad spectrum of clinical manifestations, mainly involving the nervous system, the eye and skin, and a predisposition to develop multiple benign and malignant neoplasms. Ocular diagnostic hallmarks of NF1 include optic gliomas, iris Lisch nodules, orbital and eyelid neurofibromas, eyelid café-au-lait spots. Choroidal nodules and microvascular abnormalities have recently been identified as additional NF1-related ocular manifestations. The present study was designed to describe the features and clinical significance of a new sign related to the visual apparatus in NF-1, represented by hyperpigmented spots (HSs) of the fundus oculi.ResultsHSs were detected in 60 (24.1%) out of 249 patients with NF1, with a positive predictive value of 100% and a negative predictive value of 44.2%. None of the healthy subjects (150 subjects) showed the presence of HSs. HSs were visible under indirect ophthalmoscopy, ultra-wide field (UWF) pseudocolor imaging and red-only laser image, near-infrared reflectance (NIR)-OCT, but they were not appreciable on UWF green reflectance. The location and features of pigmentary lesions matched with the already studied NF1-related choroidal nodules. No significant difference was found between the group of patients (n = 60) with ocular HSs and the group of patients (n = 189) without ocular pigmented spots in terms of age, gender or severity grading of the disease. A statistically significant association was demonstrated between the presence of HSs and neurofibromas (p = 0.047), and between the presence of HSs and NF1-related retinal microvascular abnormalities (p = 0.017).ConclusionsWe described a new ocular sign represented by HSs of the fundus in NF1. The presence of HSs was not a negative prognostic factor of the disease. Following multimodal imaging, we demonstrated that HSs and choroidal nodules were consistent with the same type of lesion, and simple indirect ophthalmoscopy allowed for screening of HSs in NF1.

Cochlear signal alterations using pseudo-color perceptual enhancement for patients with sensorineural hearing loss.

Neuroimaging detection of sensorineural hearing loss (SNHL)-related temporal bone abnormalities is limited (20-50%). We hypothesize that cochlear signal differences in gray-scale data may exceed the threshold of human eye detection. Gray-scale images can be post-processed to enhance perception of tonal difference using "pseudo-color" schemes. To compare patients with unilateral SNHL to age-matched normal magnetic resonance imaging (MRI) exams for "labyrinthine color differences" employing pseudo-color post-processing. The MRI database at an academic children's hospital was queried for "hearing loss." Only unilateral SNHL cases were analyzed. Sixty-nine imaging exams were reviewed. Thirteen age-matched normal MR exams in children without hearing loss were chosen for comparison. Pseudo-color was applied with post-processing assignment of specific hues to each gray-scale intensity value. Gray-scale and pseudo-color images were qualitatively evaluated for signal asymmetries by a board-certified neuroradiologist blinded to the side of SNHL. Twenty-six SNHL (mean: 7.6±3years) and 13 normal control exams (mean: 7.3±4years) were included. All patients had normal gray-scale cochlear signal and all controls had symmetrical pseudo-color signal. However, pseudo-color images revealed occult asymmetries localizing to the SNHL ear with lower values in 38%. Ninety-one percent of these cases showed concordance between the side of pseudo-color positivity and the side of hearing loss. Pseudo-color perceptual image enhancement reveals intra-labyrinthine fluid alterations on MR exams in children with unilateral SNHL. Pseudo-color image enhancement techniques improve detection of cochlear pathology and could have therapeutic implications.

Restoration method of sootiness mural images based on dark channel prior and Retinex by bilateral filter

Environmental changes and human activities can cause serious degradation of murals, where sootiness is one of the most common problems of ancient Chinese indoor murals. In order to improve the visual quality of the murals, a restoration method is proposed for sootiness murals based on dark channel prior and Retinex by bilateral filter using hyperspectral imaging technology. First, radiometric correction and denoising through band clipping and minimum noise fraction rotation forward and inverse transform were applied to the hyperspectral data of the sootiness mural to produce its denoised reflectance image. Second, a near-infrared band was selected from the reflectance image and combined with the green and blue visible bands to produce a pseudo color image for the subsequent sootiness removal processing. The near-infrared band is selected because it is better penetrating the sootiness layer to a certain extent comparing to other bands. Third, the sootiness covered on the pseudo color image was preliminarily removed by using the method of dark channel prior and by adjusting the brightness of the image. Finally, the Retinex by bilateral filter was performed on the image to get the final restored image, where the sootiness was removed. The results show that the images restored by the proposed method are superior in variance, average gradient, information entropy and gray scale contrast comparing to the results from the traditional methods of homomorphic filtering and Gaussian stretching. The results also show the highest score in comprehensive evaluation of edges, hue and structure; thus, the method proposed can support more potential studies or sootiness removal in real mural paintings with more detailed information. The method proposed shows strong evidence that it can effectively reduce the influence of sootiness on the moral images with more details that can reveal the original appearance of the mural and improve its visual quality.

Study on the Diagnostic Value of Magnetic Resonance Imaging in Patients with Malignant Glioma

Objective: To explore the value of multi-voxel-based monomer 1H-MRS (magnetic resonance imaging) in the classification of gliomas. Methods: 26 gliomas confirmed by surgery and pathology were examined by conventional MRI, multi-voxel 2D, and monomer 1H-MRS before surgery. Three types of regions of interest (ROI) were used to measure Cho/NAA (Choline hydroxide solution/N-acetyl aspartic acid) values in the tumor. Methods: In the multi-voxel 2D hydrogen proton spectrum, the solid part of the tumor or the most intensive region of the tumor was selected to determine the Cho/NAA value. At the same time, in the multi-voxel 2D hydrogen proton spectrum, the Fun/tool spectra's copy-2DBmin analysis software was used to obtain the Cho/NAA value distribution. Anatomically superimposed the pseudo-color image, and measure the Cho/NAA value in its largest Cho/NAA value region. In addition, based on the level set algorithm, it uses the Otsu algorithm to initially locate the edema area and provide the initial contour. Results: The Cho/NAA values measured by the ROI selection method were statistically significant, and the sensitivity and specificity of judging the benign and malignant gliomas indicated that Method 3 had the highest diagnostic value. Conclusion: Multi-voxel-based monomeric 1H-MRS magnetic resonance imaging has the highest value in guiding glioma classification and can improve the accuracy of glioma classification.