Image Acquisition Process Research Articles

Neural network technology adaptation to the small-size objects identification in satellite images of insufficient resolution within the graphic reference images database

A novel flowchart for small-size objects identification in satellite images of insufficient resolution within the graphic reference images database using neural network technology based on compromise contradiction, i.e. simultaneously the resolution enhancement of the object segment of input image and the resolution reduction of the reference image to joint resolution through the simulation of the imaging system has been proposed. This is necessary due to a significant discrepancy between the resolutions of the input image and the graphic reference images used for identification. The required level of resolution enhancement for satellite images, as a rule, is unattainable, and a significant coarsening of reference images is undesirable because of identification errors. Therefore, a certain intermediate spatial resolution is used for identification, which, on the one hand, can be obtained, and on the other the loss of information contained in the reference image is still acceptable. The intermediate resolution is determined by simulating the process of image acquisition with satellite imaging system. To facilitate such simulation, it is advisable to perform it in the frequency domain, where the advanced Fourier analysis is available and, as a rule, all the necessary transfer properties of the links of image formation chain are known. Three main functional elements are engaged for identification: an artificial neural network for the resolution enhancement of input images, a module of frequency-domain simulating of the graphical reference satellite imaging and an artificial neural network for comparing the enhanced object segment with the reference model images. The feasibility of the described approach is demonstrated by the example of successful identification of the sea vessel image in the SPOT-7 satellite image. Currently, the works are under way to compare the performance of a neural network platforms variety for small-size objects identification in satellite images aa well as to assess achievable accuracy.

Evaluating Feature Extraction Methods with Synthetic Noise Patterns for Image-Based Modelling of Texture-Less Objects

Image-based three-dimensional (3D) reconstruction is a process of extracting 3D information from an object or entire scene while using low-cost vision sensors. A structure-from-motion coupled with multi-view stereo (SFM-MVS) pipeline is a widely used technique that allows 3D reconstruction from a collection of unordered images. The SFM-MVS pipeline typically comprises different processing steps, including feature extraction and feature matching, which provide the basis for automatic 3D reconstruction. However, surfaces with poor visual texture (repetitive, monotone, etc.) challenge the feature extraction and matching stage and affect the quality of reconstruction. The projection of image patterns while using a video projector during the image acquisition process is a well-known technique that has been shown to be successful for such surfaces. In this study, we evaluate the performance of different feature extraction methods on texture-less surfaces with the application of synthetically generated noise patterns (images). Seven state-of-the-art feature extraction methods (HARRIS, Shi-Tomasi, MSER, SIFT, SURF, KAZE, and BRISK) are evaluated on problematic surfaces in two experimental phases. In the first phase, the 3D reconstruction of real and virtual planar surfaces evaluates image patterns while using all feature extraction methods, where the patterns with uniform histograms have the most suitable morphological features. The best performing pattern from Phase One is used in Phase Two experiments in order to recreate a polygonal model of a 3D printed object using all of the feature extraction methods. The KAZE algorithm achieved the lowest standard deviation and mean distance values of 0.0635 mm and −0.00921 mm, respectively.

CNN based approach for identifying banana species from fruits

Classification and recognition of fruits are still the challenging one as there are different classes of fruit types having wide inter-class resemblance. This paper proposes a banana identification model using a five-layer convolution neural network (CNN) which composed of convolution layer, pooling layer and fully connected layer. Fruits like Apple, Strawberry, and Orange, Mango and banana have been analyzed and several features have been extracted using deep learning-based CNN algorithm. Finally, the fruit identification process is done by random forest and K-Nearest Neighborhood (K-NN) classifying algorithms. A regular digital camera is used for the image acquisition process. All image manipulation processes are performed in MATLAB-17 environment.Experiments are conducted in our database consisting of 5887 fruit images. The performance of the proposed deep learning based-random forest and KNN classifiers are compared with the existing feature extraction method of HOG based random forest and KNN in terms of accuracy, precision, recall and f-score and we have achieved a accuracy rate of 96.98% for deep feature-random forest classification combination algorithm which is better than the deep feature-KNN, HOG based KNN and random forest classifiers.

Curved neuromorphic image sensor array using a MoS2-organic heterostructure inspired by the human visual recognition system

Conventional imaging and recognition systems require an extensive amount of data storage, pre-processing, and chip-to-chip communications as well as aberration-proof light focusing with multiple lenses for recognizing an object from massive optical inputs. This is because separate chips (i.e., flat image sensor array, memory device, and CPU) in conjunction with complicated optics should capture, store, and process massive image information independently. In contrast, human vision employs a highly efficient imaging and recognition process. Here, inspired by the human visual recognition system, we present a novel imaging device for efficient image acquisition and data pre-processing by conferring the neuromorphic data processing function on a curved image sensor array. The curved neuromorphic image sensor array is based on a heterostructure of MoS2 and poly(1,3,5-trimethyl-1,3,5-trivinyl cyclotrisiloxane). The curved neuromorphic image sensor array features photon-triggered synaptic plasticity owing to its quasi-linear time-dependent photocurrent generation and prolonged photocurrent decay, originated from charge trapping in the MoS2-organic vertical stack. The curved neuromorphic image sensor array integrated with a plano-convex lens derives a pre-processed image from a set of noisy optical inputs without redundant data storage, processing, and communications as well as without complex optics. The proposed imaging device can substantially improve efficiency of the image acquisition and recognition process, a step forward to the next generation machine vision.

Adjusting fictitious domain parameters for fairly priced image-based modeling: Application to the regularization of Digital Image Correlation

The integration of numerical simulation and experimental measurements in cellular materials at the sub-cellular scale is a real challenge. On the experimental side, the almost absence of texture makes displacement fields measurement tricky. On the simulation side, it requires the construction of reliable and specimen-specific geometric and mechanical models from digital images. For this purpose, high order based fictitious domain approaches have proven to be an efficient alternative to boundary conforming finite elements for the analysis of geometrically complex objects. A number of discretization parameters needs to be set by the user by making a trade-off between accuracy and computational cost. In addition to numerical errors (interpolation, integration etc.), there are additional geometric and model errors due to the pixelation of the image (e.g., quantization, sampling, noise). In the literature, discretization parameters are often analyzed without taking pixelation into account, which can lead to over-calculations. In this paper, these parameters are adjusted to obtain (a) the best possible accuracy (bounded by pixelation errors) while (b) ensuring minimal complexity (concept of fair price). In order to analyze the different sources of error, various two-dimensional synthetic experiments are generated by mimicking the image acquisition process from high-resolution numerical simulations considered as a reference. The approach leads to a modeling that outperforms conventional approaches both in terms of accuracy and complexity. Eventually, it is shown that the presented image-based models provide a unique opportunity to assist digital volume correlation and allow the measurement of relevant local kinematics within cellular materials.

Mamographic image for breast cancer detection and identification of stages of cancer using MFFC and optimized ANFIS

In women, breast cancer (BC) is basically the most often diagnosed cancers. Using mammography, the BC can be diagnosed at a premature stage, and this is the standard technique. The existing paper has posited different mammogram image detection with disparate algorithms; but their accuracy percentage is too low for a medical images since these images are highly affected by different sorts of noise. In addition, only some existing works could recognize the cancer stages. To trounce these issues, a mammographic image for BC detection and identification of stages of cancer utilizing MFFC and OANFIS is proposed. Initially, the image acquisition (IA) process is done on the strength of the morphological opening in addition to closing operation. Then, the adaptive medians filter (AMF) removes the noise as of the mammogram image. Afterward, the edge is preserved; additionally, the contrast is ameliorated as of the noise-eradicated image utilizing anisotropic diffusion histogram equalization (ADHE). Next, the mapping based shifting silhouette (MSS) removes the pectoral part of the contrast-enhanced image. Subsequently, the modified farthest first clustering (MFFC) algorithm clusters the image into (1) normal and (2) abnormal. Then, the Improved region growing (IRG) segment the tumor part from the abnormal image. Later, the features are extracted as of the segmented tumor image. Lastly, as of the extracted features, the OANFIS identifies the tumor stage. The experiments are carried out for analyzing the proposed methods’ performance. The proposed clustering as well as classification algorithms is weighted against the prevailing algorithms concerning some performance measures. The proposed MFFC along with OANFIS exhibits its superiority by attaining the highest accuracy of 99.07 and 99.63%.

Nerve Fascicle Depiction at MR Microscopy and High-Frequency US with Anatomic Verification.

In correlation to histologic cross-sections, nerve microanatomy depiction is possible with high-frequency US; however, MR microscopy is superior to high-frequency US as even finer anatomic details ...

CSR-Net: Camera Spectral Response Network for Dimensionality Reduction and Classification in Hyperspectral Imagery

Hyperspectral image (HSI) classification has become one of the most significant tasks in the field of hyperspectral analysis. However, classifying each pixel in HSI accurately is challenging due to the curse of dimensionality and limited training samples. In this paper, we present an HSI classification architecture called camera spectral response network (CSR-Net), which can learn the optimal camera spectral response (CSR) function for HSI classification problems and effectively reduce the spectral dimensions of HSI. Specifically, we design a convolutional layer to simulate the capturing process of cameras, which learns the optimal CSR function for HSI classification. Then, spectral and spatial features are further extracted by spectral and spatial attention modules. On one hand, the learned CSR can be implemented physically and directly used to capture scenes, which makes the image acquisition process more convenient. On the other hand, compared with ordinary HSIs, we only need images with far fewer bands, without sacrificing the classification precision and avoiding the curse of dimensionality. The experimental results of four popular public hyperspectral datasets show that our method, with only a few image bands, outperforms state-of-the-art HSI classification methods which utilize the full spectral bands of images.

A novel [formula omitted] norm-based superposed SR recognition framework via adaptive weighted dictionary learning

Sparse representation (SR) has played a critical role in the field of biometric image recognition for its strong robustness to noise interference. An important condition for making SR-based methods work is that the training data are totally uncorrupted. However, it may not be guaranteed in practical applications since the image acquisition process could not be carried out in a completely ideal circumstance. This paper aims to address this challenging issue by performing adaptive dictionary learning in the training stage. Following the ℓp norm and adaptive weighted Schatten-p regularization, the learned dictionary is decomposed into a low-rank component, a sparse component, and a dense noise component, where the low-rank one stands for the intrinsic class-specific structure textures, the sparse one stands for the sparse intra-class salient variations, and the noise one stands for the interference information in the corrupted training data, respectively. Furthermore, to facilitate the proposed dictionary learning framework, an improved superposed SR-based classification method is proposed by employing the ℓp norm regularization, which achieves a sparser representation compared with the ℓ1 norm regularization. This effectively enhances the representation precision since the noise interference is separately approximately represented by the decomposed dictionary. Finally, the proposed method is applied in the field of biometric image recognition, such as the palmprint recognition and the face recognition, to validate its effectiveness and improvement. Extensive experimental results on benchmark datasets show that the proposed method can effectively enhance the noise robustness and improve the recognition accuracy compared with some state-of-the-art methods.

FFA-DMRI: A Network Based on Feature Fusion and Attention Mechanism for Brain MRI Denoising.

Magnetic Resonance Imaging (MRI) is an indispensable tool in the diagnosis of brain diseases due to painlessness and safety. Nevertheless, Rician noise is inevitably injected during the image acquisition process, which leads to poor observation and interferes with the treatment. Owing to the complexity of Rician noise, using the elimination method of Gaussian to remove it does not perform well. Therefore, the feature fusion and attention network (FFA-DMRI) is proposed to separate noise from observed MRI. Inspired by the attention-guided CNN network (ADNet) and Convolutional block attention module (CBAM), a spatial attention mechanism has been specially designed to obtain the area of interest in MRI. Furthermore, the feature fusion block concatenates local with global information, which makes full use of the multilevel structure and boosts the expressive ability of network. The comprehensive experiments on Alzheimer’s disease neuroimaging initiative dataset (ADNI) have demonstrated high effectiveness of FFA-DMRI with maintaining the crucial brain details. Moreover, in terms of visual inspections, the denoising results are also consistent with human perception.

Multi-wavelength imaging of HPTLC plates using a constructed illumination chamber with a smartphone camera as the detector

In this article, we present a new type of inexpensive multi-illumination source chamber. The innovation makes use of a smartphone camera which possesses the ability of capturing multiple images. Its performance was compared to a commercially available densitometer. Similar devices and suitable strategies for data analysis will help to solve diverse classification and/or regression problems, which will be far beyond a TLC characterization of ink samples. The multi-illumination chamber was used in an exemplary forensic application. The differences in the chemical composition of various brands of fountain pen inks were revealed on images of high-performance thin-layer chromatographic plates. Reducing image data simplified the visualization and facilitated a multivariate exploratory of the ink samples. Compared to the samples that were characterized by single wavelength densitograms, the multi-wavelength characterization using the illumination chamber with a smartphone camera or densitometer improved the clustering tendency of studied samples and enhanced their interpretation. The constructed chamber for multi-wavelength imaging is an inexpensive alternative (ca. 20 Euros) to the commercially available densitometers. The discussed approaches for image acquisition and chemometric data processing support a more reliable and objective analysis of TLC multi-wavelength data.

Wind Turbine Blade Defect Image Acquisition System

The detection of wind turbine blades in service is a difficult problem in wind power operation and maintenance. Wind farms in the quality assurance unit and daily inspection, regular detection of blade is with the aid of SLR camera, then the human recognition, low accuracy and low recognition efficiency. This paper puts forward a set wind turbine blade detection method based on based on artificial intelligence image recognition system. With the help of high definition camera array and image recognition software, a large number of image acquisition and rapid recognition processing are carried out for wind turbine blades in service, and finally the automatic detection of blade surface defects is realized. The feasibility and applicability of this method are verified through field detection, which can be promoted in the field of wind power.

Critical care ultrasonography during COVID-19 pandemic: The ORACLE protocol.

Coronavirus disease 2019 (COVID-19) is characterized by severe lung involvement and hemodynamic alterations. Critical care ultrasonography is vital because it provides real time information for diagnosis and treatment. Suggested protocols for image acquisition and measurements have not yet been evaluated. This cross-sectional study was conducted at two centers from 1 April 2020 to 30 May 2020 in adult patients with confirmed COVID-19 infection admitted to the critical care unit. Cardiac and pulmonary evaluations were performed using the ORACLE protocol, specifically designed for this study, to ensure a structured process of image acquisition and limit staff exposure to the infection. Eighty-two consecutively admitted patients were evaluated. Most of the patients were males, with a median age of 56years, and the most frequent comorbidities were hypertension and type 2 diabetes, and 25% of the patients had severe acute respiratory distress syndrome. The most frequent ultrasonographic findings were elevated pulmonary artery systolic pressure (69.5%), E/e' ratio>14 (29.3%), and right ventricular dilatation (28%) and dysfunction (26.8%). A high rate of fluid responsiveness (82.9%) was observed. The median score (19 points) on pulmonary ultrasound did not reveal any variation between the groups. Elevated pulmonary artery systolic pressure was associated with higher in-hospital mortality. The ORACLE protocol was a feasible, rapid, and safe bedside tool for hemodynamic and respiratory evaluation of patients with COVID-19. Further studies should be performed on the alteration in pulmonary hemodynamics and right ventricular function and its relationship with outcomes.

Technical note: A digital technique and platform for assessing dairy cow teat-end condition

Because infections with pathogenic bacteria entering the mammary gland through the teat canal are the most common cause of mastitis in dairy cows, sustaining the integrity of the teat canal and its adjacent tissues is critical to resist infection. The ability to monitor teat tissue condition is therefore a key prerequisite for udder health management in dairy cows. However, to date, routine assessment of teat-end condition is limited to cow-side visual inspection, making the evaluation a time-consuming and expensive process. Here, we illustrate and demonstrate a method for assessing teat-end condition of dairy cows through digital images and software. A digital workflow has been designed where images of dairy cow teats are obtained and processed to display individual teats, and the cow and teat images are labeled and displayed through a graphical user interface. The interface then allows an evaluator to assess quarter- and cow-level teat-end condition and store the results for review and future analysis. The digital workflow permits several advantages such as the ability to perform remote teat-end condition assessments, and assess inter- and intrarater variability of teat-end condition scoring. We demonstrate the image-based teat-end condition assessment of 194 dairy cows that also had cow-side teat-end condition assessments by 2 expert evaluators. Weighted Cohen's kappa statistic (κ) was computed to measure the evaluators' concordance of categorical scores of quarter- and cow-level assessments when using cow-side and image-based assessments. Substantial agreement (0.61 ≤ κ ≤ 0.80) was observed between an evaluator's cow-side and image-based assessments at the quarter and cow level. Moderate agreement (0.41 ≤ κ ≤ 0.60) was observed between evaluators when using image-based assessments at the quarter and cow level. Near perfect agreement (κ = 0.89, 95% confidence interval 0.78-1.00) was observed between evaluators when using cow-side assessments at the quarter level, and substantial agreement (κ = 0.66, 95% confidence interval 0.53-0.79) was observed when using cow-side assessments at the cow level. This suggests that image-based teat-end condition classification is possible, and coupled with improvements in image acquisition and image processing, this method can be used to assess teat-end condition in a systematic and convenient manner.

Abstract 1448: Cell DIVE™: A robust and standardized platform for multiplexed whole slide imaging and single cell analysis

Abstract Multiplexed tissue imaging provides unprecedented and novel insights into cell populations and spatial interactions. Cell DIVE™ multiplexed immunofluorescence (MxIF) platform (GEHC, Issaquah, WA) provides an iterative staining, imaging and dye inactivation workflow for over 60 biomarkers on a single fixed tissue section. It is also relatively unique in providing a robust calibration process (allowing multiple calibrated instruments to produce equivalent results) using a multi-function calibration plate. Automated software enables system calibrations and robust automated image processing (field flattening, autofluorescence (AF) removal, image registration). Calibration is particularly important when imaging whole tissue slide sections where 100s of fields of view (FOV) may need to be imaged (depending on sample dimensions) and stitched for down-stream visualization and single cell analysis. In the current study, we applied our calibrated whole slide imaging process to whole tissue tonsil sections, a commonly used control for immune and cancer biomarkers. Sections were stained with 30 antibodies conjugated with either Cy3 or Cy5 dyes, including DAPI, cytokeratin AE1, CD3, CD4, CD8, CD20, CD31, CD163, COX2, FOXP3, HIF1A, HLA1, K167, LAG3, MHCII, NAKATPASE, PCAD, PCK26, S6, S100 and VIMENTIN. Each antibody underwent a robust multi-step validation process to ensure staining sensitivity and specificity. After a two-step antigen retrieval process, background autofluorescence (AF) imaging and staining, samples were imaged in the DAPI, Cy3 and Cy 5 channels on a calibrated IN Cell 2500 with Cell DIVE image acquisition and image processing software. Over 850 fields of view (FOV) were imaged in each round, followed by AF removal, registered with baseline DAPI, and stitched with minimal “quilting” artefacts. Pseudo-colored virtual H&E were also generated for the entire stitched image using DAPI and AF to assist histology interpretation. Images were imported into Halo® image analysis software (www.indicalab.com) for single cell segmentation and phenotype quantification for all FOVs. Using a subset of markers, the germinal center, non-germinal center and squamous epithelium were segmented using a random forest classifier. The HighPlex FL module was used to segment and count the total number of cells in each region, and combinations of immune cell markers were used to determine phenotype. In summary we demonstrate a standardized staining, imaging and image processing workflow for whole slide tonsil images that can be analyzed at region and cell level. Multiplexed whole slide imaging allows deep contextual analysis of cancer tissue and provides new insights into tissue heterogeneity, cell-to-cell interactions and immune profile. Citation Format: Alex Corwin, Elizabeth McDonough, Christine Surrette, Prashanth Dumpuri, Anup Sood, Fiona M. Ginty. Cell DIVE™: A robust and standardized platform for multiplexed whole slide imaging and single cell analysis [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1448.

SPECKLE DENOISING FILTERS – A SURVEY

Medical images are most commonly corrupted by various types of noises during the process of image acquisition and transmission. It is very important to get precise images to obtain exact information for the particular application. Removal of Noise from medical images is remarkably hard without knowing the prior information of noise types, causes, and removal procedures. This becomes the main reason to analyze the various noises removing process. Until now numerous researchers have working to remove noise using spatial, adaptive and fuzzy filtering approaches. This paper presents a study of various denoising procedures available in the literature.

Linear and Non-Linear Models for Remotely-Sensed Hyperspectral Image Visualization

The visualization of hyperspectral images still constitutes an open question and may have an important impact on the consequent analysis tasks. The existing techniques fall mainly in the following categories: band selection, PCA-based approaches, linear approaches, approaches based on digital image processing techniques and machine/deep learning methods. In this article, we propose the usage of a linear model for color formation, to emulate the image acquisition process by a digital color camera. We show how the choice of spectral sensitivity curves has an impact on the visualization of hyperspectral images as RGB color images. In addition, we propose a non-linear model based on an artificial neural network. We objectively assess the impact and the intrinsic quality of the hyperspectral image visualization from the point of view of the amount of information and complexity: (i) in order to objectively quantify the amount of information present in the image, we use the color entropy as a metric; (ii) for the evaluation of the complexity of the scene we employ the color fractal dimension, as an indication of detail and texture characteristics of the image. For comparison, we use several state-of-the-art visualization techniques. We present experimental results on visualization using both the linear and non-linear color formation models, in comparison with four other methods and report on the superiority of the proposed non-linear model.

Fully convolutional networks for void segmentation in X-ray images of solder joints

Whether in sensory or illumination applications, optoelectronic components are an essential part of our everyday life. To actually include them in smartphones, cars or other products, they need to be soldered onto the surface of a printed circuit board via surface-mounted technology. Hereby, the solder joint is formed by remelting the solder paste that was printed onto the board before the optical component was mounted. During this process, the evaporating flux causes porosities (voids) filled with gas that is caught in the solidifying alloy. Voids influence the thermal and electric properties of a solder joint and hence reduce its reliability. Because the solder joint is embedded between the device and the board, non-destructive X-ray inspection is used to visualize voids. However, the superposition of various structures in a noisy image acquisition process renders the semantic segmentation of solder joints and voids in X-ray images difficult. To the best of our knowledge, there is no method for automatic void segmentation in flat solder joints based on X-ray images aside from this work. We develop a fully convolutional network for pixel-wise classification of X-ray images and show, how our contributions enable automatic void inspection of soldered structures without protracted X-ray tomography of flat samples, so-called laminography.

Microscope Hardware and Software Delays Cause Photo-Toxicity

Abstract:

High efficiency dense light field and all-in-focus compression for lossless satellite image by using CCSDS

Light-field enables us to observe scenes from free view-points. However, it generally consists of 4-D enormous data, which are not suitable for storing or transmitting without effective compression. Image acquisition, transmission, interpretation and visualization have witnessed a lot of research interest and have manifold applications in real life. Light field Images popularly known as Plenoptic Images are light rays travelling in every direction. The data sets generated using plenoptic images are large in size and consume a lot of storage space. This paper studies the general process of image acquisition, light field images and various datasets. Image compression will help the users or professionals to not only achieve storage efficiency but also aid in transmission of Images. In this paper, we study efficient compression of multi-focus images synthesized from dense light-field by using DWT instead of DCT-based compression in order to suppress degradation such as block noise. Quality of reconstructed light-field is evaluated by PSNR for analyzing characteristics of residuals. This paper presents the modeling, design and implementation of two IP cores that are compliant with the CCSDS 121.0-B-2 and CCSDS 123.0-B-1 lossless satellite image compression standards. This paper provides a comparative study of various image compression techniques/algorithms and evaluates them on parameters like peak signal to noise ratio (PSNR), mean square error (MSE), Structural Similarity Index (SSI) and compression ratio (CR).