Entire Image Plane Research Articles

Stand-Alone Composite Attention Network for Concrete Structural Defect Classification

Automation in structural health monitoring involves a critical step of automatic classification of concrete defect images/videos. Although interdisciplinary research community in AI has responded with some progress, immense challenges are still involved because of the predominant overlapping nature of the defect classes, exacerbated by the large variations in their visual appearance. However, current methodologies mostly consider single-class nonoverlapping defects and emphasize equally over the entire image plane; thus unable to focus on specific defect regions for robust feature selection. Thereby, the classification performance gets degraded and the methodologies became less suitable for real-world scenarios. In this work, we propose a novel stand-alone composite attention network that automatically exerts higher emphasis on the defective regions and less emphasis on the healthy regions to recognize multitarget multiclass and single-class concrete structural defects. This architecture stems from the novel GFGA mechanism as the building block to capture minute local features from visually-similar defect classes. We then propose the MSAM that encompasses multiscale discriminative information to capture variations in image properties. The MSAM incorporates BMAM to obtain crucial channel-spatial descriptors making the overall architecture an end-to-end-trainable network. Extensive experimentation and analysis on three large concrete defect datasets show the superiority of our proposed network as compared to the current state-of-the-art methodologies.

Single Image Dehazing Using Haze-Lines

Haze often limits visibility and reduces contrast in outdoor images. The degradation varies spatially since it depends on the objects' distances from the camera. This dependency is expressed in the transmission coefficients, which control the attenuation. Restoring the scene radiance from a single image is a highly ill-posed problem, and thus requires using an image prior. Contrary to methods that use patch-based image priors, we propose an algorithm based on a non-local prior. The algorithm relies on the assumption that colors of a haze-free image are well approximated by a few hundred distinct colors, which form tight clusters in RGB space. Our key observation is that pixels in a given cluster are often non-local, i.e., spread over the entire image plane and located at different distances from the camera. In the presence of haze these varying distances translate to different transmission coefficients. Therefore, each color cluster in the clear image becomes a line in RGB space, that we term a haze-line. Using these haze-lines, our algorithm recovers the atmospheric light, the distance map and the haze-free image. The algorithm has linear complexity, requires no training, and performs well on a wide variety of images compared to other state-of-the-art methods.

Visibility enhancement of images degraded by hazy weather conditions using modified non-local approach

Natural images captured under bad weather situations suffer from poor visibility and contrast problems. Object tracking and recognition under hazy bad weather conditions is a very difficult task for real time applications. Therefore, in this paper, we have proposed an efficient dehazy algorithm for visibility and contrast enhancement of color hazy images. The proposed algorithm works in two phases. In the first phase, a non local approach is applied in the hazy model, which is a pixel based approach not a patch based. Pixels are spread over the entire image plane and positioned at different distance from the sensor, so this approach is called non local. Degradation is different for every pixel therefore; estimation of the transmission map for every pixel through the haze line is the essential step. After the first phase, the image becomes unnatural and dimmed, therefore to proper tone mapping and improving the visual quality of the image, we applied the S-shaped mapping function in the second phase. The quantitative results of the proposed algorithm and other existing dehazy algorithms for color hazy images are obtained in terms of Hazy Reduction factor(HRF), and measure of enhancement factor(EMF) on different hazy image databases. Qualitative results reveal that the visual quality of the proposed algorithm is better than other existing de-hazy algorithms. Simulation results demonstrate that the proposed algorithm provides better results as compared to other existing dehazy algorithms for color hazy images. Proposed algorithm is highly efficient as compare to other latest dehazy algorithms.

High-accuracy spectral reduction algorithm for the échelle spectrometer.

A spectral reduction algorithm for an échelle spectrometer with spherical mirrors that builds a one-to-one correspondence between the wavelength and pixel position is proposed. The algorithm accuracy is improved by calculating the offset distance of the principal ray from the center of the image plane in the two-dimensional vertical direction and compensating the spectral line bending from the reflecting prism. The simulation and experimental results verify that the maximum deviation of the entire image plane is less than one pixel. This algorithm ensures that the wavelengths calculated from spectrograms have a high spectral resolution, meaning the precision from the spectral analysis reaches engineering standards of practice.

Low-Aberration Imaging Monochromator with Dispersion Subtraction Based on an Axially Symmetric Design

An axially symmetric optical system for an imaging monochromator with dispersion subtraction for multiple multispectral scans is proposed and simulated numerically. This system provides minimal angles of incidence for beams on spherical objectives so that off-axis aberrations are reduced over the entire image plane. The width of the selected spectral band for individual image segments is independent of the entrance aperture size. A method of eliminating background illumination of the detector is proposed and the light losses are estimated.

ON THE SOURCE OF ASTROMETRIC ANOMALOUS REFRACTION

More than a century ago, astronomers using transit telescopes to determine precise stellar positions were hampered by an unexplained periodic shifting of the stars they were observing. With the advent of CCD transit telescopes in the past three decades, this unexplained motion, termed "anomalous refraction" by these early astronomers, is again being observed. Anomalous refraction is described as a low-frequency, large angular scale (∼2°) motion of the entire image plane with respect to the celestial coordinate system as observed and defined by astrometric catalogs. These motions, of typically several tenths of an arcsecond amplitude with timescales on the order of 10 minutes, are ubiquitous to ground-based drift-scan astrometric measurements regardless of location or telescopes used and have been attributed to the effect of tilting of equal-density layers of the atmosphere. The cause of this tilting has often been attributed to atmospheric gravity waves, but this cause has never been confirmed. Although theoretical models of atmospheric refraction show that atmospheric gravity waves are a plausible cause of anomalous refraction, an observational campaign specifically directed at defining this relationship provides clear evidence that anomalous refraction is not consistent with the passage of atmospheric gravity waves. The source of anomalous refraction is found to be meter-scale, slowly evolving quasi-coherent dynamical structures in the boundary layer below 60 m above ground level.

Swept Field Laser Confocal Microscopy for Enhanced Spatial and Temporal Resolution in Live-Cell Imaging

Confocal fluorescence microscopy is a broadly used imaging technique that enhances the signal-to-noise ratio by removing out of focal plane fluorescence. Confocal microscopes come with a variety of modifications depending on the particular experimental goals. Microscopes, illumination pathways, and light collection were originally focused upon obtaining the highest resolution image possible, typically on fixed tissue. More recently, live-cell confocal imaging has gained importance. Since measured signals are often rapid or transient, thus requiring higher sampling rates, specializations are included to enhance spatial and temporal resolution while maintaining tissue viability. Thus, a balance between image quality, temporal resolution, and tissue viability is needed. A subtype of confocal imaging, termed swept field confocal (SFC) microscopy, can image live cells at high rates while maintaining confocality. SFC systems can use a pinhole array to obtain high spatial resolution, similar to spinning disc systems. In addition, SFC imaging can achieve faster rates by using a slit to sweep the light across the entire image plane, thus requiring a single scan to generate an image. Coupled to a high-speed charge-coupled device camera and a laser illumination source, images can be obtained at greater than 1,000 frames per second while maintaining confocality.

Parametric Perfusion Imaging Based on Low-Cost Ultrasound Platform

In this study, we attempted to implement parametric perfusion imaging to quantify blood perfusion based on modified low-cost ultrasound platform. A novel ultrasound contrast-specific imaging method called pulse-inversion harmonic sum-squared-differences (PIHSSD) was proposed for improving the sensitivity for detecting contrast agents and the accuracy of parametric perfusion imaging, which combined pulse-inversion harmonic (PIH) with pulse-inversion sum-squared-differences (PISSD) threshold-based decision. PIHSSD method just involved simple operations including addition and multiplication and was easy to realize. The sequences of contrast images without logarithmic compression were used to acquire time intensity curves (TICs) from numerous equal-sized regions-of-interest (ROI) covering the entire image plane. Parametric perfusion images were obtained based on the parameters extracted from the TICs, including peak value (PV), area under curve (AUC), mean transit time (MTT), peak value time (PVT), peak width (PW) and climbing rate (CR). Flow phantom was used for validation and the results suggested that PIHSSD method provided 9.6 to 20.3 dB higher contrast-to-tissue ratio (CTR) than PIH method. The results of the experiments of rabbit kidney also showed that the CTR of PIHSSD images was higher than that of PIH images, and the parametric perfusion images based on PIHSSD method provided more accurate quantification of blood perfusion compared with those based on PIH and PISSD methods. It demonstrated that the parametric perfusion imaging achieved good performance though implemented on low-cost ultrasound platform. (E-mail: mxwan@mail.xjtu.edu.cn)

An efficient region of interest acquisition method for dynamic magnetic resonance imaging

Motivated by work in the area of dynamic magnetic resonance imaging (MRI), we develop a new approach to the problem of reduced-order MRI acquisition. Efforts in this field have concentrated on the use of Fourier and singular value decomposition (SVD) methods to obtain low-order representations of an entire image plane. We augment this work to the case of imaging an arbitrarily-shaped region of interest (ROI) embedded within the full image. After developing a natural error metric for this problem, we show that determining the minimal order required to meet a prescribed error level is in general intractable, but can be solved under certain assumptions. We then develop an optimization approach to the related problem of minimizing the error for a given order. Finally, we demonstrate the utility of this approach and its advantages over existing Fourier and SVD methods on a number of MRI images.

A study on particle identification in PTV particle mask correlation method

An algorithm for particle identification in PTV, which is classified in the category .particle mask correlation method,.is presented and tested. A typical brightness pattern of a particle image is referred to as a particle mask. The particle mask is centered on a pixel in the image plane and the cross correlation between the particle mask and a region of the image plane centered on the given pixel is calculated. The particle mask is scanned over the entire image plane and the cross correlation coefficient is calculated at each pixel location. Small subareas with high correlation coefficients indicate central areas of particle images.

Parallel attentive visual tracking

The research reported here addresses the problem of detecting and tracking independently moving objects from a moving observer in real-time, using corners as object tokens. Corners are detected using the Harris corner detector, and local image-plane constraints are employed to solve the correspondence problem. The approach relaxes the restrictive static-world assumption conventionally made, and is therefore capable of tracking independently moving and deformable objects. Tracking is performed without the use of any 3-dimensional motion model. The technique is novel in that, unlike traditional feature-tracking algorithms where feature detection and tracking is carried out over the entire image-plane, here it is restricted to those areas most likely to contain-meaningful image structure. Two distinct types of instantiation regions are identified, these being the “focus-of-expansion” region and “border” regions of the image-plane. The size and location of these regions are defined from a combination of odometry information and a limited knowledge of the operating scenario. The algorithms developed have been tested on real image sequences taken from typical driving scenarios. Implementation of the algorithm using T800 Transputers has shown that near-linear speedups are achievable, and that real-time operation is possible (half-video rate has been achieved using 30 processing elements).

The analytic structure of image flows: Deformation and segmentation

Time-varying imagery is often described in terms of image flow fields (i.e., image motion), which correspond to the perceptive projection of feature motions in three dimensions (3D). In the case of multiple moving objects with smooth surfaces, the image flow possesses an analytic structure that reflects these 3D properties. This paper describes the analytic structure of image flow fields in the image space-time domain, and its use for segmentation and 3D motion computation. First we discuss the local flow structure as embodied in the concept of neighborhood deformation. The local image deformation is effectively represented by a set of 12 basis deformations, each of which is responsible for an independent deformation. This local representation provides us with sufficient information for the recovery of 3D object structure and motion, in the case of relative rigid body motions. We next discuss the global flow structure embodied in the partitioning of the entire image plane into analytic regions separated by boundaries of analyticity, such that each small neighborhood within the analytic region is described in terms of deformation bases. This analysis reveals an effective mechanism for detecting the analytic boundaries of flow fields, thereby segmenting the image into meaningful regions. The notion of consistency which is often used in the image segmentation is made explicit by the mathematical notion of analyticity derived from the projection relation of 3D object motion. The concept of flow analyticity is then extended to the temporal domain, suggesting a more robust algorithm for recovering image flow from multiple frames. Finally, we argue that the process of flow segmentation can be understood in the framework of grouping process. The general concept of coherence or grouping through local support (such as the second-order flows in our case) is discussed.

Fraunhofer diffraction by semicircular apertures

The far-field diffraction pattern of semicircular apertures is described. A general expression is given for the amplitude that is valid over the entire image plane and that reduces to Bessel and Struve functions on the axes. A simple approximation is also given that is valid for all angles at large distances from the origin.