Pixel Distance Research Articles

Distance Measurement Using Dual Laser Source and Image Processing Techniques

Camera based systems are gaining popularity in automotive driver assistance systems. They are used to detect obstacles, road signs, road markings and other road users. Today, even low cost camera technology enables obtaining a very good image resolution. The proposed system presented in this paper can be used to measure the distance or the displacement of an object, by applying laser optics and trigonometry. The vision-based system was developed using a dual laser source and a CCD camera for recording images which are processed to obtain the distance from an object. The image processing techniques used in this paper consists of two main improved processes which are, the detection of a laser spot using image processing and the calculation of the distance based on laser spot position on the image and the relationship between pixel number and distance. It is shown that the proposed system is able to measure distances in a quick and accurate fashion.

Improving random walker segmentation using a nonlocal bipartite graph

The classical random walk segmentation explores merely local affinity among neighboring pixels for cutting out objects, which falls short of effectiveness when handling distant repetitive patterns. Meanwhile, the running efficiency is also limited by solving a large scale linear system. To alleviate the quandary, in this paper, we first propose to introduce nonlocal affinity among distant pixels with similar local features in the underlying segmentation graph, which enables label propagation among disconnected foregrounds and thus multiple repetitive patterns can be segmented jointly. Secondly, the segmentation graph is extended to a multi-scale superpixel based bipartite graph, where random walker segmentation is performed to transfer labeling information from annotated seeds to superpixels and further to unlabeled pixels. We show such a bipartite graph based approach can considerably save the computational cost without sacrificing the segmentation accuracy. Extensive experiments are conducted on multiple datasets, demonstrating the effectiveness of the proposed method.

Machine Learning Model based Image and Video Colorization

In this paper, we present an intuitive framework for clients to effectively colorize the regular pictures of complex scenes. In our framework, colorization system is expressly isolated into two phases: Shading naming and Shading planning. Pixels that ought to generally have comparative tones are assembled into lucid districts in the shading naming stage, and the shading planning stage is then acquainted with additional tweak the tones in each intelligible locale. To deal with textures ordinarily found in normal pictures, we propose another shading marking plan that bunches not just adjoining pixels with comparative force yet additionally distant pixels with comparable surface. Propelled by the knowledge into the complementary nature moved by the exceptionally contrastive areas and the smooth areas, we utilize a perfection guide to direct the joining of power progression and surface comparability limitations in the plan of our labeling calculation. Inside each sound area acquired from the shading marking stage, the shading planning is applied to produce distinctive colorization impact by allotting shadings to a couple of pixels in the district. A bunch of natural interface apparatuses is intended for marking, shading and changing the outcome. We show convincing consequences of colorizing assuming measure of client input.

Quantitative distribution characterization of gradient composition of additive-manufactured stainless steel using micro-beam X-ray fluorescence

The fabrication of the compositionally graded metals provides a new frontier in additive manufacturing and the characterization of the elemental distribution of graded materials is strongly desired for the quality control of powder mixing and improvement of additive manufacturing process parameters. In this study, Micro-beam X-ray fluorescence (μ-XRF) was used to characterize the elemental distribution of additive-manufactured compositionally graded stainless steel on a large area. A poly-capillary X-ray optics unit was used to obtain a high spatial resolution with the spot size of 20-μm. The quantitative distribution characterization of Ni, Mo, Cr, Mn, Si and Cu elements in the whole deposition direction have been realized through the improvement of μ-XRF quantitative calibration method by using certified reference materials with composition and microstructure similar to the graded sample. The composition distribution variation in different cladded layer was also investigated with the pixel distance decreased to 20 μm. Moreover, laser-induced breakdown spectroscopy was used to analyze the composition distribution of the composition-graded steels to verify the reliability of quantitative distribution characterization results by μ-XRF method. The elemental distribution results from two methods have good agreement, but the results determined by μ-XRF can give more details about the composition distribution in different cladded layers because of its higher spatial resolution, which is closely related to additive manufacturing process. The microstructure content varied linearly with the change of composition. The austenite volume fraction increased with an increase in Ni content and the evolution rule of the structure agreed with the calculated thermodynamics model results. Cracks appeared in the martensite and austenite transformation region with a large volume change when the Ni content ranged between 6.5% and 8.5% for the gradient stainless steels.

Fixation stability and deviation in optical coherence tomography angiography using soft contact lens correction in myopes

To compare fixation deviation and stability with soft contact lens correction and device built-in auto-focus system during optical coherence tomography angiography (OCTA). This observational study measured OCTA metrics first with contact lens correction, followed by removal of contact lenses, using the device auto-focus system at a University Optometry Clinic, Hong Kong. All participants were habitual soft contact lens wearers with either low or high myopia. OCTA measurements were obtained using a spectral domain OCTA. Fixation deviation was distance (in pixels) of the fovea to the center of the OCTA measurement grid. Fixation stability was test–retest repeatability (TRR) and coefficient of variation (CV) of fixation deviation from three consecutive OCTA measurements. OCTA metrics included vessel length density (VD), perfusion density (PD), and foveal avascular zone (FAZ) area. Averaged OCTA metrics were calculated from three measurements and compared between the two correction methods. The mean ± SD spherical equivalent of 74 eyes from 74 myopes measured was − 1.94D ± 0.75D in low myopes (n = 37) and − 7.97D ± 1.31D in high myopes (n = 37). When corrected with contact lenses, visual acuities of high myopes (median [IQR], − 0.06 [0.08] logMAR) and low myopes (− 0.02 ± 0.08 logMAR) were similar (P = 0.060), and with similar fixation deviation (5.0 ± 2.2 pixels vs 5.3 [3.6] pixels; P = 0.689). High myopes had poorer fixation stability than low myopes (TRR: 10.2 pixels vs 7.5 pixels; CV: 65% vs 54%, respectively). The worst fixation stability occurred when high myopes were corrected using the auto-focus system (TRR: 12.5 pixels, CV: 72%). The difference in VD and PD was within 1 mm−1 and 1%, respectively. The FAZ area was similar. Difference in OCTA metrics was small in each refractive group (< 1 mm−1 in VD, and < 2% in PD). High myopes had more stable fixation when corrected when wearing contact lenses. Subjects with good contact lens corrected visual acuity should wear their contact lenses during OCTA measurements.

Echocardiography image enhancement using texture-cartoon separation

Due to the speckled nature of cardiac ultrasound imaging, it is not easy to process and extract useful information directly from the acquired image. In this work, we have proposed a method to reduce the effect of speckle artifacts through the decomposition of echocardiography images into cartoon and texture components. The first component (i.e., cartoon image) contains image structures containing smooth areas and sharp edges, and the texture component is mainly composed of highly oscillating and repetitive patterns. To decompose the image into these two subcomponents, convolutional sparse coding has been utilized as a solid tool for solving the decomposition optimization function. The significant advantage of using convolutional sparse coding, compared to classical sparse coding methods, is image quality enhancement due to not using the block coding, making the classic solutions computationally feasible.The original image has been masked with the cartoon part leading to suppress speckle artifacts which result in image quality enhancement. Besides, it has been shown that using this speckle reduction scenario, considerable accuracy enhancement of the segmentation task can be achieved, compared to segmentation of the original image. Numerical results provide acceptable reasons to prove the efficiency of the proposed algorithm. Resulting echocardiography videos show a mean segmentation enhancement of 15.98 for Hausdorff distance (in pixels) and 0.0632 for the Dice similarity coefficient.

Object-Gaze Distance: Quantifying Near- Peripheral Gaze Behavior in Real-World Applications.

Eye tracking (ET) has shown to reveal the wearer’s cognitive processes using the measurement of the central point of foveal vision. However, traditional ET evaluation methods have not been able to take into account the wearers’ use of the peripheral field of vision. We propose an algorithmic enhancement to a state-of-the-art ET analysis method, the Object- Gaze Distance (OGD), which additionally allows the quantification of near-peripheral gaze behavior in complex real-world environments. The algorithm uses machine learning for area of interest (AOI) detection and computes the minimal 2D Euclidean pixel distance to the gaze point, creating a continuous gaze-based time-series. Based on an evaluation of two AOIs in a real surgical procedure, the results show that a considerable increase of interpretable fixation data from 23.8 % to 78.3 % of AOI screw and from 4.5 % to 67.2 % of AOI screwdriver was achieved, when incorporating the near-peripheral field of vision. Additionally, the evaluation of a multi-OGD time series representation has shown the potential to reveal novel gaze patterns, which may provide a more accurate depiction of human gaze behavior in multi-object environments.

Structure-Consistent Weakly Supervised Salient Object Detection with Local Saliency Coherence

Sparse labels have been attracting much attention in recent years. However, the performance gap between weakly supervised and fully supervised salient object detection methods is huge, and most previous weakly supervised works adopt complex training methods with many bells and whistles. In this work, we propose a one-round end-to-end training approach for weakly supervised salient object detection via scribble annotations without pre/post-processing operations or extra supervision data. Since scribble labels fail to offer detailed salient regions, we propose a local coherence loss to propagate the labels to unlabeled regions based on image features and pixel distance, so as to predict integral salient regions with complete object structures. We design a saliency structure consistency loss as self-consistent mechanism to ensure consistent saliency maps are predicted with different scales of the same image as input, which could be viewed as a regularization technique to enhance the model generalization ability. Additionally, we design an aggregation module (AGGM) to better integrate high-level features, low-level features and global context information for the decoder to aggregate various information. Extensive experiments show that our method achieves a new state-of-the-art performance on six benchmarks (e.g. for the ECSSD dataset: Fβ = 0.8995, Eξ = 0.9079 and MAE = 0.0489), with an average gain of 4.60% for F-measure, 2.05% for E-measure and 1.88% for MAE over the previous best performing method on this task. Source code is available at http://github.com/siyueyu/SCWSSOD.

Explicitly Modeled Attention Maps for Image Classification

Self-attention networks have shown remarkable progress in computer vision tasks such as image classification. The main benefit of the self-attention mechanism is the ability to capture long-range feature interactions in attention-maps. However, the computation of attention-maps requires a learnable key, query, and positional encoding, whose usage is often not intuitive and computationally expensive. To mitigate this problem, we propose a novel self-attention module with explicitly modeled attention-maps using only a single learnable parameter for low computational overhead. The design of explicitly modeled attention-maps using geometric prior is based on the observation that the spatial context for a given pixel within an image is mostly dominated by its neighbors, while more distant pixels have a minor contribution. Concretely, the attention-maps are parametrized via simple functions (e.g., Gaussian kernel) with a learnable radius, which is modeled independently of the input content. Our evaluation shows that our method achieves an accuracy improvement of up to 2.2% over the ResNet-baselines in ImageNet ILSVRC and outperforms other self-attention methods such as AA-ResNet152 in accuracy by 0.9% with 6.4% fewer parameters and 6.7% fewer GFLOPs. This result empirically indicates the value of incorporating geometric prior into self-attention mechanism when applied in image classification.

Looking Wider for Better Adaptive Representation in Few-Shot Learning

Building a good feature space is essential for the metric-based few-shot algorithms to recognize a novel class with only a few samples. The feature space is often built by Convolutional Neural Networks (CNNs). However, CNNs primarily focus on local information with the limited receptive field, and the global information generated by distant pixels is not well used. Meanwhile, having a global understanding of the current task and focusing on distinct regions of the same sample for different queries are important for the few-shot classification. To tackle these problems, we propose the Cross Non-Local Neural Network (CNL) for capturing the long-range dependency of the samples and the current task. CNL extracts the task-specific and context-aware features dynamically by strengthening the features of the sample at a position via aggregating information from all positions of itself and the current task. To reduce losing important information, we maximize the mutual information between the original and refined features as a constraint. Moreover, we add a task-specific scaling to deal with multi-scale and task-specific features extracted by CNL. We conduct extensive experiments for validating our proposed algorithm, which achieves new state-of-the-art performances on two public benchmarks.

Cherry Tree Crown Extraction from Natural Orchard Images with Complex Backgrounds

Highly effective pesticide applications require a continual adjustment of the pesticide spray flow rate that attends to different canopy characterizations. Real-time image processing with rapid target detection and data-processing technologies is vital for precision pesticide application. However, the extant studies do not provide an efficient and reliable method of extracting individual trees with irregular tree-crown shapes and complicated backgrounds. This paper on our study proposes a Mahalanobis distance and conditional random field (CRF)-based segmentation model to extract cherry trees accurately in a natural orchard environment. This study computed Mahalanobis distance from the image’s color, brightness and location features to acquire an initial classification of the canopy and background. A CRF was then created by using the Mahalanobis distance calculations as unary potential energy and the Gaussian kernel function based on the image color and pixels distance as binary potential energy. Finally, the study completed image segmentation using mean-field approximation. The results show that the proposed method displays a higher accuracy rate than the traditional algorithms K-means and GrabCut algorithms and lower labeling and training costs than the deep learning algorithm DeepLabv3+, with 92.1%, 94.5% and 93.3% of the average P, R and F1-score, respectively. Moreover, experiments on datasets with different overlap conditions and image acquisition times, as well as in different years and seasons, show that this method performs well under complex background conditions, with an average F1-score higher than 87.7%.

Development of an automatic measurement device for double laser shield tail gap based on image recognition technology

Shield tail clearance is an important basis for the selection of segment assembly and shield attitude control. It is the most direct factor affecting shield tail brush wear. Poor shield tail clearance is prone to segment damage and shield tail seal failure. Using manual measurement of the shield tail gap, the measurement data has a certain hysteresis, and can not directly reflect the changing trend of the shield tail gap. Based on image recognition technology, a dual laser shield tail gap measurement device was developed to measure the changes in the shield tail gap during shield tunneling. The field test shows that: 1 The measurement device can monitor the shield tail gap in real time, and can simulate the ellipticity of the shield tail, monitor the deformation of the shield tail and the space state of the shield tail and the tube. 2 The acquisition device adopts a dual laser transmitter, meanwhile, the attitude of the shield machine cannot be controlled in real time, Can accurately establish the geometric relationship between the actual distance and the pixel distance, improve the test accuracy of the measuring device, and control the monitoring accuracy at 3.8 mm.

A Tree Ring Measurement Method Based on Error Correction in Digital Image of Stem Analysis Disk

Stem analysis is an essential aspect in forestry investigation and forest management, as it is a primary method to study the growth law of trees. Stem analysis requires measuring the width and number of tree rings to ensure the accurate measurement, expand applicable tree species, and reduce operation cost. This study explores the use of Open Source Computer Vision Library (Open CV) to measure the ring radius of analytic wood disk digital images, and establish a regression equation of ring radius based on image geometric distortion correction. Here, a digital camera was used to photograph the stem disks’ tree rings to obtain digital images. The images were preprocessed with Open CV to measure the disk’s annual ring radius. The error correction model based on the least-square polynomial fitting method was established for digital image geometric distortion correction. Finally, a regression equation for tree ring radius based on the error correction model was established. Through the above steps, click the intersection point between the radius line and each ring to get the pixel distance from the ring to the pith, then the size of ring radius can be calculated by the regression equation of ring radius. The study’s method was used to measure the digital image of the Chinese fir stem disk and compare it with the actual value. The results showed that the maximum error of this method was 0.15 cm, the average error was 0.04 cm, and the average detection accuracy reached 99.34%, which met the requirements for measuring the tree ring radius by stem disk analysis. This method is simple, accurate, and suitable for coniferous and broad-leaved species, which allows researchers to analyze tree ring radius measurement, and is of great significance for analyzing the tree growth process.

A Study on The Minimum Exposure Dose For Image Formation In X-Ray Imaging of Extremity

Excellent performance of the digital detector can generate medical images at a lower dose. However, if a certain level of dose exposure of the sensor detector can generate an image to recognize it.&#x0D; In this study, the exposure conditions of the forearm, femur, and ankle joint tests currently used in clinical trials were tested with lower doses. The resolution pattern was analyzed using the Image J program, and the results were obtained by analyzing the Interactive 3D Surface Plot. In addition, by measuring the surface dose and absorbed dose, the digital detector finds the minimum conditions to make an image, and also checks the effect of dose reduction.&#x0D; The image of the obtained resolution pattern was analyzed by Plot profile using Image J program. It was confirmed that the gray value width decreased at the same pixel distance as the exposure conditions decreased. Even if the exposure conditions were lowered from step 1 to step 4, which is the standard condition of the forearm test, the resolution showed no difference at 1.4 Lp/mm.&#x0D; In the Interactive 3D Surface Plot, as the range of the gray scale lowered the exposure condition, the shape of the surface contours gradually became blurred and the contrast of black and white shades decreased. The resolution from step 1 to step 3, which is the standard condition of the femur bone test, was 1.2 Lp/mm, indicating no difference. Even if the exposure conditions were lowered from step 1 to step 4, which is the standard condition for ankle examination, the resolution showed no difference at 1.4 Lp/mm.&#x0D; It is thought that it is necessary to accumulate a lot of data by further subdividing the stage of exposure conditions according to the thickness in more areas.

Using Self-Aligned Si Barrier to Enhance the Contrast Ratio and Performance of Pixelated Light Emitting Diode for Vehicle Headlamp

We investigate the effects of pixel distance and light emitting diode (LED) thickness on the luminance distribution and contrast ratio of a 16 × 16 blue and white LED array for vehicle headlamp application. With increasing pixel distance from 10 to 20 μm, the contrast ratio increased from 33:1 to 51:1 for white LEDs and from 37:1 to 48:1 for blue LEDs. As the LED thickness decreased from 4.5 to 2.0 μm, the contrast ratio increased from 37:1 to 62:1 for blue LEDs and from 33:1 to 97:1 for white LEDs. The blue LEDs fabricated with ultrathin-GaN and self-aligned Si barrier showed the contrast ratio of 48:1 and 39:1, respectively, while the white LEDs with ultrathin-GaN and self-aligned Si barrier yielded the contrast ratio of and 89:1 and 99:1. Further, LEDs with the self-aligned Si barrier had better luminance uniformity than those with the ultrathin-GaN. LED array with the self-aligned Si barrier exhibited narrower dark space than that with the ultrathin-GaN. At 60 mA, the LED with the self-aligned Si barrier gave 6% higher light output than that with the ultrathin-GaN. A headlamp unit using 16 × 16 LED array was demonstrated with the Si self-aligned barrier.

Newly designed identifying method for ice thickness on high‐voltage transmission lines via machine vision

The icing of transmission lines will bring considerable challenges to the safe operation of the power grid. Therefore, a novel method combines machine vision and machine learning algorithms for identifying the ice thickness on high-voltage transmission line (HVTL) as proposed herein. First, noise and background interference in the image are filtered, and the grey image is used as input. Then, the algorithms of improved Canny edge detection, Hough transform, improved K-means clustering, and least-squares fitting are adopted in turn to locate the edges of conductors. Finally, according to the distance mapping model based on monocular vision, the ice thickness of the conductor is determined by calculating the width difference before and after icing. The experimental results show that the proposed method can accurately locate the edge of the conductor in both field and experimental environments. Moreover, it can ensure ideal effects under different illumination and hardly not be affected by distortion in both horizontal and vertical directions. Besides, the distance mapping model can map the pixel distance to the actual distance with high precision, no matter whether the background is simple or complex, and the calculated ice thickness has only a small deviation compared to the actual value. In addition, the proposed method shows high reliability and effectiveness when various interference such as different backgrounds, uneven icing, height difference changes, conductor movement, contrast changes, and conductor sag occur.

Three-dimensional evaluation method for asphalt pavement texture characteristics

To supplement the research on the evaluation method of asphalt pavement texture, novel three-dimensional (3D) methods are proposed. First, 18 different pavement textures were measured in laboratory from asphalt mixture specimens using laser texture scanner (LTS), and the macro-texture and micro-texture were extracted based on spectrum analysis techniques. Then, macro-texture level evaluation indices f8mac and f9mac together with micro-texture level evaluation indices f8mic and f9mic were proposed based on the gray level co-occurrence matrix (GLCM) method, and the determination of hyperparameters angle θ, pixel distance d and vertical resolution v were discussed. Through the correlation analysis and compared with existing conventional evaluation indices, the optimum pavement texture level evaluation index was selected and verified. Additionally, the evaluation index σ of distribution uniformity of pavement texture (DUPT) was proposed based on the uniformity of deviations between sub-surfaces and the average surface of pavement texture. Finally, the correlations of σ with texture profiles were studied. The results show that f8mac and f8mic are determined as the optimum indices for pavement texture level. Mean texture depth has significant correlation with f8mac, and the correlation coefficient R is 0.9611; friction coefficient μ has significant correlation with f8mic, and the R is 0.9422. θ is recommended to be 90°; d and v are set to 52 pixels (0.4953 mm) and 0.01 mm respectively for macro-texture level evaluation; d and v are set to 1 pixel (0.0095 mm) and 0.003 mm respectively for micro-texture level evaluation. Moreover, the effectiveness of σ is also validated by calibrating with standard grooved surface. It can be concluded that the proposed indices in this study are suitable to the evaluation of pavement texture level and pavement texture distribution.

Significance driven inverse distance weighted filter to restore impulsive noise corrupted X-ray image

This paper presents a novel significance driven inverse distance weighted (SDIDW) filter for the impulsive noise removal in the X-ray images. The proposed SDIDW filter restores the noisy pixel using minimum number of nearest noise-free pixels to achieve good estimation while exhibiting low computational complexity. In the proposed filter, higher priority (weight) is given to nearest pixels compared to distant pixels and only sufficient nearest noise free pixels are determined to estimate the value of noisy pixel. A high level analysis of the computation complexity at varying noise density is done which shows that proposed SDIDW filter provides significant reduction in computation complexity over the adaptive median filters. Finally, the performance of the proposed filter is evaluated and compared over the state-of-the-art impulse noise removal techniques for varying noise density (wide range 10–90% and very high noise density range 91–99%). The experimental results on medical images demonstrate significant improvement in filtered images quality by the proposed filter over the state-of-the-art filters at each sample of noise density with small computational complexity.

RETRACTED: Design and implementation of parallel algorithm for image matching based on Hausdorff Distance

Object matching two-dimensional images in computer vision has become a significant subject of article acknowledgment and picture investigation. Hausdorff Distance assumes a significant function in coordinating image. To proposed system parallel algorithm for image matching manage the instance of arbitrary commotion, image coordinating, new Hausdorff Distance is proposed in this. In contrast to coordinating two twofold pictures, different techniques, the proposed strategy might be coordinated with a few dark scale image pixel esteems. One case of article acknowledgment is utilized to show the productivity of the proposed strategy. The outcomes demonstrated that, contrasted and, the new Hausdorff Distance (HD) might be more alluring approach to discard image clamor coordinating, because of the extensive reflection to decide the dark scale data Hausdorff Distance of adjoining pixels in the shooting of his realities into account. Besides, the strategy can be acknowledged in a straightforward manner.

Reversible Data Hiding Based on Adaptive Multiple Histograms Modification

Pixel value ordering prediction has been verified as an effective mechanism to exploit image redundancy for reversible data hiding (RDH) and numerous extensions have been devised. However, their performance is still unsatisfactory since the error modification is generally fixed and independent of image content. In this paper, a new RDH scheme is proposed by incorporating pixel distance to realize adaptive multiple histograms modification (AMHM). During exploiting the correlation between the largest/smallest pixel and any other one in the scope of pixel block, we propose to process every two correlated pixels successively following the ascending order of their distance. Specifically, the generated errors with a given distance are collected and verified. If they are all shiftable errors, the follow-up errors would be collected into the next sub-histogram. In this way, a histogram sequence is adaptively generated such that different modification mechanisms can be taken for different sub-histograms to achieve adaptive embedding. Finally, AMHM for conventional prediction-error expansion (PEE) and AMHM for 2D PEE have been both realized in this paper. Experimental results show that AMHM is of great significance to better exploit pixel correlation and the proposed scheme outperforms a series of the latest schemes.