Multi-scale Gradient Research Articles

A Novel Image Fusion Method for Multi-focus Image Fusion Based on the Nonsubsampled Contourlet Transform and Multi-scale Statistic

A novel image fusion method based on the Nonsubsampled Contourlet Transform (NSCT) and multiscale statistic for multi-focus image fusion is proposed. The objective of multi-focus image fusion is to combine the clear part of different source images with the same scene to a single image, and how to get a better focus measurement is the key to the question. The paper managed to analysis the inputs image in NSCT domain with multi-scale statistical features and gives a novel fusion strategy to select the coefficients. The multi-scale average gradient is applied on the low frequency coefficients. And all of the high frequency coefficients of the same position were selected from the same source image decided by the average of the multi-scale energy value of these subbands. The comparative result with the other methods based on Wavelet Transform (WT), Contourlet Transform (CT) and NSCT shows that the proposed method exacts the most information from the source images and achieves better performance both in the subjective visual perception and quantitative metrics.

Nanometer-thick lateral polyelectrolyte micropatterns induce macrosopic electro-osmotic chaotic fluid instabilities

Electro-convective vortices in ion concentration polarization under shear flow have been of practical relevance for desalination processes using electrodialysis. The phenomenon has been scientifically disregarded for decades, but is recently embraced by a growing fluid dynamics community due its complex superposition of multi-scale gradients in electrochemical potential and space charge interacting with emerging complex fluid momentum gradients. While the visualization, quantification and fundamental understanding of the often-chaotic fluid dynamics is evolving rapidly due to sophisticated simulations and experimentation, little is known whether these instabilities can be induced and affected by chemical topological heterogeneity in surface properties. In this letter, we report that polyelectrolyte layers applied as micropatterns on ion exchange membranes induce and facilitate the electro-osmotic fluid instabilities. The findings stimulate a variety of fundamental questions comparable to the complexity of today's turbulence research.

A marker-based watershed method for X-ray image segmentation

Digital X-ray images are the most frequent modality for both screening and diagnosis in hospitals. To facilitate subsequent analysis such as quantification and computer aided diagnosis (CAD), it is desirable to exclude image background. A marker-based watershed segmentation method was proposed to segment background of X-ray images. The method consisted of six modules: image preprocessing, gradient computation, marker extraction, watershed segmentation from markers, region merging and background extraction. One hundred clinical direct radiograph X-ray images were used to validate the method. Manual thresholding and multiscale gradient based watershed method were implemented for comparison. The proposed method yielded a dice coefficient of 0.964±0.069, which was better than that of the manual thresholding (0.937±0.119) and that of multiscale gradient based watershed method (0.942±0.098). Special means were adopted to decrease the computational cost, including getting rid of few pixels with highest grayscale via percentile, calculation of gradient magnitude through simple operations, decreasing the number of markers by appropriate thresholding, and merging regions based on simple grayscale statistics. As a result, the processing time was at most 6s even for a 3072×3072 image on a Pentium 4 PC with 2.4GHz CPU (4 cores) and 2G RAM, which was more than one time faster than that of the multiscale gradient based watershed method. The proposed method could be a potential tool for diagnosis and quantification of X-ray images.

Demosaicing of Captured Images using Multiscale Gradient and Heterogeneity Projection

Color images require multiple data samples for each pixel as opposed to grayscale images for which a pixel is represented by only one data sample. For RGB format, these data samples represent red, green, blue channels. A typical digital camera captures only one of these channels at each pixel location and the other need to be estimated to generate complete color information This process is known as color filter array interpolation (CFA).The objective of the proposed research is to develop high performance, low computational complexity resolution enhancement and demosaicing algorithms. Our approach to both problems is to end creative ways to incorporate edge information into the algorithm design. However, in contrast with the usual edge directed approaches, we do not try to detect edge presence and orientation explicitly. For the image interpolation problem, we study the relationship between low resolution and high resolution pixels, and derive a general interpolation formula to be used on all pixels. This simple interpolation algorithm is able to generate sharp edges in any orientation. Additionally, we propose a gradient based directional demosaicing method that does not require setting any thresholds. We show that the performance of this algorithm can be improved by using multiscale gradients. Finally, we address the low spectral correlation Demosaicing problem by proposing a new family of hybrid colour filter array (CFA) patterns and a local algorithm that is two orders of magnitude faster than a comparable non-local solution while offering the same level of performance.

Splitting touching cells based on concave-point and improved watershed algorithms

Splitting touching cells is important for medical image processing and analysis system. In this paper, a novel strategy is proposed to separate ellipse-like or circle-like touching cells in which different algorithms are used according to the concave-point cases of touching domains. In the strategy, a concave-point extraction and contour segmentation methods for cells in series and in parallel are used for the images with distinct concave points, and an improved watershed algorithm with multi-scale gradient and distance transformation is adopted for the images with un-distinct or complex concave points. In order to visualize each whole cell, ellipse fitting is used to process the segments. Experimental results show that, for the cell images with distinct concave points, both of the two algorithms can achieve good separating results, but the concave-point based algorithm is more efficient. However, for the cell images with unobvious or complex concave points, the improved watershed based algorithm can give satisfying segmenting results.

Multiscale shear-strain gradient for detecting delamination in composite laminates

Delamination is a typical failure mode in composite laminates. For detecting delamination, a new physical concept of a multiscale shear-strain gradient (MSG) is formulated from a vibrational mode shape of a Kirchhoff plate. In a laminate containing a small delamination, the MSG characterizes shear-strain singularities with an explicit physical implication and no need for material properties, in turn manifesting the distribution of the delamination; moreover, multiscale property enables the MSG to identify delaminations at a high noise level. Numerical and experimental results show that the MSG defies the traditional understanding that vibrational features are inappropriate to localize a slight delamination.

Multi-Scale Descriptor for Robust and Fast Camera Motion Estimation

This letter presents a robust and fast 6D camera motion estimation system that combines a FAST detector with a Multi-scale Descriptor (MSD). For a robust and fast camera motion estimation system, the descriptor is an especially important component, as it is the most time consuming step in motion estimation. Furthermore, the properties of the descriptor significantly affect the precision of the estimated motion. In this letter, we present a descriptor that requires a low computational burden and offers high precision for robust and fast motion estimation. The descriptor provides high precision through the use of multi-scale gradient information and computational efficiency by employing 4 integral images and an intensity gradient. We conclude this letter with an evaluation of the proposed motion estimation in comparison with different types of descriptors in practical situations. The results show that the proposed motion estimation provides good performance over previous methods in terms of processing time and precision of estimated camera motion.

Scale dependent effects in resource selection by crop-raiding Japanese macaques in Niigata Prefecture, Japan

A resource selection function is one that yields values proportional to the probability of use of a resource unit. This quantity is influenced by the heterogeneity of landscape structures, which occurs over multiple spatial scales. To provide input into wildlife management strategies, we investigated the scale dependency and functional responses of Japanese macaques using multiple scale analysis. The multiple buffers with radii of 100, 500, 1000, 1500, 2000, and 2500 m were defined as the spatial scale. Crop damage was predicted at the within-home range scale, using the Random Forests algorithm with environmental variables linked to resource selection of Japanese macaques. Sixteen environmental variables were defined, covering aspects of landscape configuration, human disturbance, topography, and adopted countermeasures. Crop damage was most accurately predicted within a buffer zone of 1000 m, although radii exceeding 1000 m were also highly accurate. Although the importance of variables differed among spatial extents, the functional responses for each environmental variable were independent of spatial extent. These results suggest that the limiting factors of crop damage depend on spatial extent, while functional responses in resource selection remain constant across spatial extents. We also compared a multi-scale gradient map with a typical binary map to demonstrate the uncertainty in damage predictions at different spatial scales. Our results may aid wildlife management planning, for which differences in resource selection across different spatial scales are critically important.

A HYBRID APPROACH FOR INFORMATION EXTRACTION FROM HIGH RESOLUTION SATELLITE IMAGERY

This paper presents a hybrid approach for extraction of information from high resolution satellite imagery and also demonstrates the accuracy achieved by the final extracted information. The hybrid technique comprises of improved marker-controlled watershed transforms and a nonlinear derivative method. It overcomes all the disadvantages of existing region-based and edge-based methods by incorporating aforesaid hybrid methods. It preserves the advantages of multi-resolution and multi-scale gradient approaches. Region-based segmentation also incorporates the watershed technique due to its better efficiency in segmentation. In principle, a proper segmentation can be performed perfectly by watershed technique on incorporating ridges. These ridges express as the object's boundaries according to the property of contour detection. On the other hand, the nonlinear derivative method is used for resolving the discrete edge detection problem. Since it automatically selects the best edge localization, which is very much useful for estimation of gradient selection. The main benefit of univocal edge localization is to provide a better direction estimation of the gradient, which helps in producing a confident edge reference map for synthetic images. The practical merit of this proposed method is to derive an impervious surface from emerging urban areas.

A Novel Model of Image Segmentation Based on Watershed Algorithm

A novel model of image segmentation based on watershed method is proposed in this paper. To prevent the oversegmentation of traditional watershed, our proposed algorithm has five stages. Firstly, the morphological reconstruction is applied to smooth the flat area and preserve the edge of the image. Secondly, multiscale morphological gradient is used to avoid the thickening and merging of the edges. Thirdly, for contrast enhancement, the top/bottom hat transformation is used. Fourthly, the morphological gradient of an image is modified by imposing regional minima at the location of both the internal and the external markers. Finally, a weighted function is used to combine the top/bottom hat transformation algorithm and the markers algorithm to get the new algorithm. The experimental results show the superiority of the new algorithm in terms of suppression over-segmentation.

MULTI-SCALE GRADIENT INVARIANT FOR FACE RECOGNITION UNDER VARYING ILLUMINATION

In this paper, a novel approach derived from image gradient domain called multi-scale gradient faces (MGF) is proposed to abstract multi-scale illumination-insensitive measure for face recognition. MGF applies multi-scale analysis on image gradient information, which can discover underlying inherent structure in images and keep the details at most while removing varying lighting. The proposed approach provides state-of-the-art performance on Extended YaleB and PIE: Recognition rates of 99.11% achieved on PIE database and 99.38% achieved on YaleB which outperforms most existing approaches. Furthermore, the experimental results on noised Yale-B validate that MGF is more robust to image noise.

Image bilevel thresholding based on stable transition region set

We utilize the linear system theory to establish a theory model of transition region. With the model, we reveal an important property of transition region, namely the gray level distribution symmetry. Utilizing the property, we propose a new thresholding framework based on stable transition region set. The elements of the stable transition region set are equal or close to each other in the average gray level. As an example of the proposed framework, we have shown that the feature transformation based on the multiscale gradient multiplication technology is an effective means of estimating the threshold. We have performed subjective and objective comparisons on both synthetic and real images. The experimental results show the segmentation quality of the proposed approach is superior to three conventional transition region-based thresholding methods.

Image bilevel thresholding based on multiscale gradient multiplication

This paper introduces a novel global thresholding approach that exploits the multiscale gradient information. The multiscale gradient information, that is, the product of gradient magnitude (PGM), is obtained by multiplying the responses of the first derivative of Gaussian (FDoG) filter at three adjacent space scales. The output threshold is selected as the one that maximizes a new objective function of the gray level variable t. The objective function is defined as the ratio of the mean PGM values of the boundary and non-boundary regions in the binary image obtained by thresholding with variable t. Through analysis of 35 real images from different application areas, our results show that the proposed method can perform bilevel thresholding on the images with different histogram patterns, such as unimodal, bimodal, multimodal, or comb-like shape. Its segmentation quality is superior to five popular thresholding algorithms.

Quantifying the tibiofemoral joint space using x‐ray tomosynthesis

Digital x-ray tomosynthesis (DTS) has the potential to provide 3D information about the knee joint in a load-bearing posture, which may improve diagnosis and monitoring of knee osteoarthritis compared with projection radiography, the current standard of care. Manually quantifying and visualizing the joint space width (JSW) from 3D tomosynthesis datasets may be challenging. This work developed a semiautomated algorithm for quantifying the 3D tibiofemoral JSW from reconstructed DTS images. The algorithm was validated through anthropomorphic phantom experiments and applied to three clinical datasets. A user-selected volume of interest within the reconstructed DTS volume was enhanced with 1D multiscale gradient kernels. The edge-enhanced volumes were divided by polarity into tibial and femoral edge maps and combined across kernel scales. A 2D connected components algorithm was performed to determine candidate tibial and femoral edges. A 2D joint space width map (JSW) was constructed to represent the 3D tibiofemoral joint space. To quantify the algorithm accuracy, an adjustable knee phantom was constructed, and eleven posterior-anterior (PA) and lateral DTS scans were acquired with the medial minimum JSW of the phantom set to 0-5 mm in 0.5 mm increments (VolumeRad™, GE Healthcare, Chalfont St. Giles, United Kingdom). The accuracy of the algorithm was quantified by comparing the minimum JSW in a region of interest in the medial compartment of the JSW map to the measured phantom setting for each trial. In addition, the algorithm was applied to DTS scans of a static knee phantom and the JSW map compared to values estimated from a manually segmented computed tomography (CT) dataset. The algorithm was also applied to three clinical DTS datasets of osteoarthritic patients. The algorithm segmented the JSW and generated a JSW map for all phantom and clinical datasets. For the adjustable phantom, the estimated minimum JSW values were plotted against the measured values for all trials. A linear fit estimated a slope of 0.887 (R² = 0.962) and a mean error across all trials of 0.34 mm for the PA phantom data. The estimated minimum JSW values for the lateral adjustable phantom acquisitions were found to have low correlation to the measured values (R² = 0.377), with a mean error of 2.13 mm. The error in the lateral adjustable-phantom datasets appeared to be caused by artifacts due to unrealistic features in the phantom bones. JSW maps generated by DTS and CT varied by a mean of 0.6 mm and 0.8 mm across the knee joint, for PA and lateral scans. The tibial and femoral edges were successfully segmented and JSW maps determined for PA and lateral clinical DTS datasets. A semiautomated method is presented for quantifying the 3D joint space in a 2D JSW map using tomosynthesis images. The proposed algorithm quantified the JSW across the knee joint to sub-millimeter accuracy for PA tomosynthesis acquisitions. Overall, the results suggest that x-ray tomosynthesis may be beneficial for diagnosing and monitoring disease progression or treatment of osteoarthritis by providing quantitative images of JSW in the load-bearing knee.

Application of Multi-Scale Morphology Gradient in the Foreign Fibers Recognition

The extraction of edge feature is a key technology in the foreign fibers recognition. The traditional algorithm of edge detection operator is more sensitive to noise, which is unable to extract complete edge information. Based on the morphological edge detection, we can get the ideal foreign fibers edge characteristics under the condition of noise for the multi-scale structure elements, combining with inflation and corrosion operator to structure morphology gradient operator, the adjustment of morphology structure elements dimensions. Experiment results show that it has a certain feasibility and effectiveness compared to classic edge detection algorithm in the respect of restraining the noise and improving the testing precision.

Tongue-coating image segmentation based on combination of morphological gradient and watershed algorithms

The segmentation of tongue-coating images is very important to computer-aided traditional Chinese medicine (TCM) clinic diagnosis system. By now, algorithms for this segmentation are in low-accuracy or high-complexity. Upon the edge-gradient feature of tongue-coating images, a new segmenting method is presented in this paper, which combines edge and region processing technologies. With the new method, the multi-scale morphological gradient images of the original images are taken firstly, and then watershed algorithm is used to segment them so that tongue-coating images are separated. The following experimental results show this method's effectiveness and efficiency.

Image Quality Assessment by Visual Gradient Similarity

A full-reference image quality assessment (IQA) model by multiscale visual gradient similarity (VGS) is presented. The VGS model adopts a three-stage approach: First, global contrast registration for each scale is applied. Then, pointwise comparison is given by multiplying the similarity of gradient direction with the similarity of gradient magnitude. Third, intrascale pooling is applied, followed by interscale pooling. Several properties of human visual systems on image gradient have been explored and incorporated into the VGS model. It has been found that Stevens' power law is also suitable for gradient magnitude. Other factors such as quality uniformity, visual detection threshold of gradient, and visual frequency sensitivity also affect subjective image quality. The optimal values of two parameters of VGS are trained with existing IQA databases, and good performance of VGS has been verified by cross validation. Experimental results show that VGS is competitive with state-of-the-art metrics in terms of prediction precision, reliability, simplicity, and low computational cost.

Information Potential Fields Navigation in Wireless Ad-Hoc Sensor Networks

As wireless sensor networks (WSNs) are increasingly being deployed in some important applications, it becomes imperative that we consider application requirements in in-network processes. We intend to use a WSN to aid information querying and navigation within a dynamic and real-time environment. We propose a novel method that relies on the heat diffusion equation to finish the navigation process conveniently and easily. From the perspective of theoretical analysis, our proposed work holds the lower constraint condition. We use multiple scales to reach the goal of accurate navigation. We present a multi-scale gradient descent method to satisfy users’ requirements in WSNs. Formula derivations and simulations show that the method is accurately and efficiently able to solve typical sensor network configuration information navigation problems. Simultaneously, the structure of heat diffusion equation allows more flexibility and adaptability in searching algorithm designs.

PATTERNED POLYMERS: Fabrication of Multiscale Gradient Polymer Patterns by Direct Molding and Spatially Controlled Reflow (Adv. Funct. Mater. 6/2011)

AbstractSize variations of pattern spacing as well as gradient control of the as‐formed polymeric pattern via a spatially controlled reflow process are presented. Micro‐ and nanopatterns of polymethyl methacrylate (PMMA) in the form of line‐and‐space strips are first generated by capillary force lithography (CFL), and the residual layers are removed by ashing process. Subsequently, the exposed PMMA strips underwent a controlled reflow process above the glass transition temperature (Tg) while heating single or both sides of the substrate either in parallel to the line pattern (parallel reflow) or perpendicular to the line pattern (perpendicular reflow). As a result of this controlled reflow, a linear or a parabolic profile of pattern spacing is achieved depending on the heating mode. Furthermore, multiscale gradient patterns are formed with the spacing ranging from 98 nm to 4.23 μm (a difference of two orders of magnitude) in a single patterned layer using the original micropattern of 16 μm width and 8 μm spacing. In order to explain reflow behaviors, a simple theoretical model relating the normalized pattern width to the polymer viscosity is derived based on a leveling kinetics of polymer melt. Also, gradient PMMA channels are fabricated and bonded to a glass substrate, which are used to flow a liquid inside the channels by capillarity‐driven flow.

基于多步长梯度特征的红外弱小目标检测算法

基于多步长梯度特征的红外弱小目标检测算法