Low-frequency Image Research Articles

High-speed low-frequency chirped coherent anti-Stokes Raman scattering microscopy using an ultra-steep long-pass filter.

Coherent anti-Stokes Raman scattering (CARS) microscopy is becoming a more common tool in biomedical research. High-speed CARS microscopy has important applications in live cell imaging and in label-free pathology. However, only a few realizations exist of CARS imaging applied in the few terahertz spectral range (<300 cm-1), in which much is unknown to date. Although single-beam CARS microscopy proved to be robust in this low-frequency region, pixel-dwell time using presently available schemes is still relatively long, in the millisecond scale. Single-beam notch-shaped chirped-CARS (C-CARS) microscopy in the fingerprint region can be performed without using lock-in detection, yet it necessitates double-notch shaping, resulting in a relatively complex system. Here, we demonstrate that C-CARS in the low-frequency regime can be achieved using a sharp-edge, which is created by an ultra-steep long-pass filter (ULPF). Furthermore, we demonstrate that this variant of C-CARS spectroscopy can be performed without post-processing analyses. This is used to image collagen in a biological sample with a pixel dwell time of 200 microseconds. This sharp-edge C-CARS method may find important application in rapid low-frequency CARS imaging of live cells or for imaging of fast flowing objects such as in microfluidic channels.

Research on hybrid fusion algorithm for multi-feature among heterogeneous image

There are many types of features in heterogeneous images, and the differences between these features are significant; thus, ensuring that different features can be fused into a single image is difficult. To address this problem, a hybrid fusion algorithm for multi-feature fusion between heterogeneous images is proposed in this paper. First, the contour features of the heterogeneous images are extracted by using the local pixel meaning value contrast, and the fusion weight of the contour feature is obtained using a guider filter with the aim of fusing the contour features between images. Second, the edge features of the heterogeneous images are extracted by using the local standard deviation, and then fused by using the local standard deviation similarity between images as the threshold, yielding the edge features of the fused image. Third, the contour-feature fusion and edge-feature fusion images are combined to obtain the low-frequency fusion image. Fourth, the source images are transformed by using two-dimensional variation mode decomposition (2D-VMD), and a detailed-feature fusion image is obtained by using the skewness and an impulse neural network (PCNN). Finally, the spatial domain fusion image is injected into 2D-VMD as the low-frequency fusion image of 2D-VMD, and the fusion result is obtained via inverse transformation of 2D-VMD. The experimental results demonstrate that the proposed fusion algorithm can clearly improve the fusion performance of heterogeneous images, and generate a more informative fusion image.

Nickel foam surface defect detection based on spatial-frequency multi-scale MB-LBP

According to the nickel foam surface defect images with the typical characteristics of complex geometry and texture distribution, a nickel foam surface defect detection method based on spatial-frequency multi-scale block local binary pattern is proposed. First, nonsubsampled contourlet is used to carry out foam nickel image multi-scale decomposition, and therefore, low-frequency sub-band images and high-frequency sub-band images are obtained. The multi-scale block local binary pattern is then used to extract the feature histogram vectors of each block region of low- and high-frequency sub-bands, and the histogram feature vectors of the whole image after cascade are formed. The kernel principal component analysis and support vector machine are adopted to reduce the dimension of the feature histogram vectors and used for the defect classification. Experimental results show that the proposed method of feature extraction can extract more detailed texture information, and the average recognition rate reaches to 90%, which meets an enterprise’s needs.

Regularization strategies for deep-learning-based salt model building

Salt model building has long been considered a severe bottleneck for large-scale 3D seismic imaging projects. It is one of the most time-consuming, labor-intensive, and difficult-to-automate processes in the entire depth imaging workflow requiring significant intervention by domain experts to manually interpret the salt bodies on noisy, low-frequency, and low-resolution seismic images at each iteration of the salt model building process. The difficulty and need for automating this task is well-recognized by the imaging community and has propelled the use of deep-learning-based convolutional neural network (CNN) architectures to carry out this task. However, significant challenges remain for reliable production-scale deployment of CNN-based methods for salt model building. This is mainly due to the poor generalization capabilities of these networks. When used on new surveys, never seen by the CNN models during the training stage, the interpretation accuracy of these models drops significantly. To remediate this key problem, we have introduced a U-shaped encoder-decoder type CNN architecture trained using a specialized regularization strategy aimed at reducing the generalization error of the network. Our regularization scheme perturbs the ground truth labels in the training set. Two different perturbations are discussed: one that randomly changes the labels of the training set, flipping salt labels to sediments and vice versa and the second that smooths the labels. We have determined that such perturbations act as a strong regularizer preventing the network from making highly confident predictions on the training set and thus reducing overfitting. An ensemble strategy is also used for test time augmentation that is shown to further improve the accuracy. The robustness of our CNN models, in terms of reduced generalization error and improved interpretation accuracy is demonstrated with real data examples from the Gulf of Mexico.

Improvement of Multi-Frequency Microwave Breast Imaging through Frequency Cycling and Tissue-Dependent Mapping

Two modifications of the Contrast Source Inversion (CSI) algorithm have been employed to improve resulting 2-D microwave images of the dielectric properties of synthetic breast models produced using a well-established frequency-hopping technique. First, while reconstructions from low-frequency data are often used as initial guesses for otherwise unstable higher-frequency inversions, an improvement in overall image quality can be observed when the imaginary part of the reconstructed low-frequency image is altered to reflect the identical tissue geometry as its real part when used as an initial guess during subsequent reconstructions. Second, in such frequency-hopping scenarios, rather than terminating reconstructions following inversions of the highest frequency data, “frequency cycling” (i.e., returning to the lowest frequency in the sequence) can be used to further improve reconstructions without loss of resolution or anatomical detail.

Nonreference Image Quality Evaluation Algorithm Based on Wavelet Convolutional Neural Network and Information Entropy

The image quality evaluation method, based on the convolutional neural network (CNN), achieved good evaluation performance. However, this method can easily lead the visual quality of image sub-blocks to change with the spatial position after the image is processed by various distortions. Consequently, the visual quality of the entire image is difficult to reflect objectively. On this basis, this study combines wavelet transform and CNN method to propose an image quality evaluation method based on wavelet CNN. The low-frequency, horizontal, vertical, and diagonal sub-band images decomposed by wavelet transform are selected as the inputs of convolution neural network. The feature information in multiple directions is extracted by convolution neural network. Then, the information entropy of each sub-band image is calculated and used as the weight of each sub-band image quality. Finally, the quality evaluation values of four sub-band images are weighted and fused to obtain the visual quality values of the entire image. Experimental results show that the proposed method gains advantage from the global and local information of the image, thereby further improving its effectiveness and generalization.

Noise detection and image denoising based on fractional calculus

Abstract Fractional-order integral can weaken high-frequency signal and greatly preserve low-frequency signal nonlinearly, that is, the high-frequency noise can be removed while retaining the information of low-frequency image itself, thus the fractional calculus can achieve good a denoising effect and the application of fractional calculus theory in the digital image processing field has also been favored by more and more scholars. On the basis of summarizing and analyzing previous research works, this paper proposes a new method for detecting noise points in images using fractional differential gradient and an improved image denoising algorithm based on fractional integration. The noise detection method determines the noise position by the fractional differential gradient, and achieves to detect the noise, snowflake and stripe anomaly though utilizing the neighborhood information feature of the image and the contour and direction distribution of various noise anomalies in spatial domain; the image denoising algorithm firstly regards the appearance of noise points in image as a small probability event and divides it, then applies the fractional calculus to subtly transform and adjust signal filter and meanwhile utilizes iterative idea to control image denoising effect for accurately distinguishing between noise and high-frequency signals, so that the original image feature information is more preserved while the image is denoised. The simulation results show that this algorithm model can effectively remove the noise while maintaining the details of image edge and texture, and have the characteristics of simple algorithm and good stability. The model can effectively remove the noise and meanwhile preserve the details of image edges and textures, and has the characteristics of simple algorithm and high stability. The study results of this paper provide a reference for further research on the noise detection and image denoising based on fractional calculus.

Sub-second hyper-spectral low-frequency vibrational imaging via impulsive Raman excitation.

Real-time vibrational microscopy has been recently demonstrated by various techniques, most of them utilizing the well-known schemes of coherent anti-stokes Raman scattering and stimulated Raman scattering. These techniques readily provide valuable chemical information mostly in the higher vibrational frequency regime (>400 cm-1). Addressing the low vibrational frequency regime (<200 cm-1) is challenging due to the usage of spectral filters that are required to isolate the signal from the Rayleigh scattered excitation field. In this Letter, we report on rapid, high-resolution, low-frequency (<130 cm-1) vibrational microscopy using impulsive coherent Raman excitation. By combining impulsive excitation with a fast acousto-optic delay line, we detect the Raman-induced optical Kerr lensing and spectral shift effects with a 25μs pixel dwell time to produce shot-noise limited, low-frequency hyper-spectral images of various samples.

Application of Edge Preserving and Interpolation Algorithm Based on Wavelet Transformation in Mechanical Inspection and Flaw Detection

In mechanical inspection and flaw detection, the fault point in the image is often small, it is easy to cause the occurrence of missed detection, and due to the internal factors of the imaging system and environmental effects, the final image is often degraded, which is manifested as the deterioration of subjective vision and the decrease of actual resolution. In order to improve the resolution of the image without changing the existing imaging system and prevent the occurrence of the phenomenon of missed detection. An interpolation edge preserving algorithm based on wavelet transformation is proposed, the algorithm firstly performs edge detection and interpolation for low-resolution image, after that, the initial interpolated magnified image was decomposed by wavelet, and the low-frequency component image after wavelet decomposition was replaced by the original low-resolution image. Then, the inverse wavelet transform was carried out to obtain the high-resolution image with good edge information. The experimental results showed that compared with the traditional interpolation method and wavelet interpolation method, this algorithm performs better in the edge visual effect and objective index, and it is an effective algorithm to improve the resolution of detection image.

A Fuzzy Approach to Recognize Face Using Contourlet Transform

Face recognition addresses identification, verification, and authentication in biometric-based security systems. This work enhances the contrast and edges in face images and recognizes the face using contourlet transform and fuzzy rules. Contourlet transformed image provides multiscale and directional information. The transformed image is divided into low-pass image (low-frequency image) and band-pass image (high-frequency image). The low-pass image is enhanced using fuzzy-based histogram specification since it deals with contrast. Band-pass image contains detailed information about the edges of the image and are enhanced using fuzzy rules and morphological gradient operators. The proposed system achieves the accuracy rate of 99.81% and 99.35% on Yale-B and JAFEE dataset, respectively, which is better than the existing curvelet and wavelet transform-based recognition. The incorporation of fuzzy rules enhances the mean intensity value of the edges to 34.19, which is better than Canny, Sobel, Prewitt, Robert and Laplacian edge detection techniques. Finally Discriminant Correlation Analysis (DCA) feature level fusion is applied to fuse enhanced edge intensities and histogram features for Support Vector Machine (SVM) classification.

Gaussian Mixture Model and Lifting Wavelet Transformed Base Satellite Image Enhancement

From the last few decades, Satellite images are being used widely in various applications like monitoring of forest areas, weather forecasting, polar bears counting, etc. In those applications to get more details of images efficiently, satellite images should be enhanced up to the required level as the images captured by the satellites are covered very large areas and those are very low-resolution images due to the high altitudes of satellites from the earth. We proposed a method of an image enhancement which includes both resolution enhancement and contrast enhancement. In this method, Stationary Wavelet Transform (SWT) in combination with Lifting Wavelet Transform (LWT) is used for image decomposition into low-frequency sub band images and high-frequency sub band images to separate smooth regions and sharp edges to interpolate regions and edges separately to reduce blurring effect in edges and noise in smooth regions. To get smoother details and sharper edges, Gaussian Mixture Model (GMM) is used for interpolation in resolution enhancement process and SWT with the combination of Contrasts Limited Adaptive Histogram Equalization (CLAHE) for contrast enhancement process. SWT in combination with LWT improves the resolution effectively and also minimizes the execution time drastically than existing methods due to the shift invariance of SWT and reduced computations in LWT and GMM interpolation results from sharper edges and smoother details. SWT is used in combination with CLAHE to enhance the contrast and mitigate the noise effects than existing methods. The proposed method gives superior results and compared with existing techniques with PSNR, Noise Estimation, and visual results.

High resolution ultrasonic imaging based on frequency sweep in both of transducer element domain and imaging line domain

We developed the super resolution FM-chirp correlation method (SCM) and its high frame rate version called synthetic aperture-SCM (SA-SCM) based on frequency sweep in order to improve the range resolution in ultrasound imaging. However, in these methods, discontinuities in the lateral direction tend to occur, because the SCM processing is performed line by line in the image. In addition, in array transducers, grating lobes are more likely to occur when the spatial resolution is increased using high-frequencies. Therefore, another version called SCM-weighted SA was proposed. In the SCM-weighted SA, in order to avoid the lateral discontinuities, first the SCM processing is performed on each echo received by each transducer element, and then the obtained SCM result is used as a weight for SA processing. The SCM-weighted SA can generate multiple B-mode images each of which corresponds to each carrier frequency, and the appropriate low frequency images among them have no grating lobes. While this plurality of B-mode images can be used as multi-frequency images, it can also be integrated as one high SNR image. In the present study, we propose a method to generate the integrated one B-mode image by applying SCM again in the line direction of the obtained multi-frequency images, which is called SCM-weighted SA-SCM. This method can further improve the range resolution.

3-D seismic exploration across the deep geothermal research platform Gro\xdf Sch\xf6nebeck north of Berlin/Germany

The North German Basin is one of the three major type localities in Germany for deep geothermal energy. Here, the pore space is the dominant parameter, in contrast to fractures (Rhine Graben) and karst (Molasse Basin). To further develop the geothermal research platform Groß Schönebeck located in the Northeast German Basin, a subset of the North German Basin, we investigated the geological structure and the existence of possible fault systems in the subsurface. For this purpose, we carried out a high-resolution 3-D reflection seismic survey at the location to overcome methodical restrictions of the few 2-D seismic profiles that cross the area of interest, such as spatial focusing effects and fault imaging. The survey area extends 8 km × 8 km at the surface and focusses down to reservoir depths at 4 km. With four vibrators as source (12–96 Hz sweep, 12 s duration), we used a source line spacing of 700 m and a receiver line spacing of 400 m with both 50 m source and geophone spacing. Data processing encompassed CRS (Common Reflection Surface) stacking, post-stack time migration and depth conversion. We observed a smooth doming of the Zechstein salt from 0.6 to 1 km thickness above the continuous top of the Rotliegend Group at around 4 km depth, and the Mesozoic horizons above appear as mainly continuous reflection surfaces with gentle undulations and occasional normal faulting. We highly resolved the supra-salt sequences in the study area for the first time, which allowed us image an almost complete suite of reflectors mapped in other parts of the Northeast German Basin. However, less resolved, lower frequency images are encountered deeper than ~ 4 km. Two important factors for further field development are that we do not observe an apparent influence of crustal-scale faults, which were expected from former conceptual models for the region, and that at the current status of work, the reservoir does not show a fracture-dominated character.

Multi-scale fusion algorithm of intensity and polarization-difference images based on edge information enhancement

To better integrate complementary and redundant information from different source images, improve the edge information, and facilitate the target detection. A multi-scale fusion algorithm of intensity and polarization-difference (PD) images based on edge information enhancement is proposed. Firstly, intensity images are obtained by the polarization information analysis method. PD images are obtained by the adaptive polarization-difference imaging approach based on the principle of minimum mutual information. Secondly, guided filter, affine transformations and Block-Matching and 3D filtering are embedded in visibility enhancement to improve the intensity and PD images. Thirdly, the two images are decomposed into high-frequency and low-frequency images by the dual-tree complex wavelet transform (DT-CWT). The high-frequency and low-frequency images are fused by the fusion rules based on edge detection and the regional variance matching degree respectively. Finally, the fusion image is obtained by the inverse DT-CWT. Experimental results demonstrate that fusion images of the proposed algorithm are significantly improved in information entropy, average gradient, and spatial frequency. Compared with the existing methods, it can achieve a better edge enhancement for images in a turbid medium.

GMRT Low-frequency Imaging of an Extended Sample of X-shaped Radio Galaxies

Abstract We present a low-frequency imaging study of an extended sample of X-shaped radio sources using the Giant Metrewave Radio Telescope (GMRT) at two frequencies (610 and 240 MHz). The sources were drawn from a Very Large Array Radio Images of the Sky at Twenty-Centimeters (FIRST) selected sample and extend an initial GMRT study at the same frequencies of 12 X-shaped radio galaxies predominantly from the 3CR catalog (Lal &amp; Rao 2007). Both the intensity maps and spectral index maps of the 16 newly observed sources are presented. With the combined sample of 28 X-shaped radio sources, we found no systematic differences in the spectral properties of the higher surface brightness, active lobes versus the lower surface brightness, or off-axis emission. The properties of the combined sample are discussed, including the possible role of a twin active galactic nuclei model in the formation of such objects.

Asymmetric color cryptosystem based on compressed sensing and equal modulus decomposition in discrete fractional random transform domain

In this paper, we propose an asymmetric encryption and compression method for a color image based on compressed sensing and equal modulus decomposition in the discrete fractional random transform (DFrRT) domain. In this scheme, a color image is firstly encrypted into a single channel image, which is divided into low-frequency image and high-frequency image by the discrete wavelet transform. Subsequently, the high-frequency image is compressed into two matrices by compressed sensing. One of the matrices is combined with the low-frequency image into one input of DFrRT. The other acts as another input of DFrRT. Finally, the output result of DFrRT is encrypted by the equal modulus decomposition in DFrRT domain to create an effective trapdoor one-way function. Compared to the other transformations, the private key of DFrRT is the high-frequency component related to the plaintext in this paper. Thus, the proposed cryptosystem is able to resist various types of attacks and maintain the asymmetric characteristics of the cryptosystem. Numerical simulations are presented to demonstrate the feasibility and robustness of the proposed method.

Ghost imaging of the low or high frequency based on the corresponding spatial-frequency of the reference pattern

We discuss the possibility that the contours or details of images can be obtained by using the low-spatial-frequency or high-spatial-frequency parts of the reference beam in ghost imaging. The influence of the threshold selection which is used to divide the reference pattern into low-frequency and high-frequency parts is quite different. With the increase of the threshold, the signal-to-noise ratio (SNR) of low-frequency ghost imaging increases firstly and then decreases, while SNR of high-frequency ghost imaging decreases gradually. Under a suitable threshold, the results are compared with those by directly filtering ghost-image from the traditional ghost imaging. It is shown that better imaging quality can be obtained by using our scheme. To further improve imaging quality, the method of subtracting background before correlation calculation is also presented.

Fast and accurate single image super-resolution via an energy-aware improved deep residual network

Recently, convolutional neural network (CNN) based single image super-resolution (SISR) solutions have demonstrated significant progress on restoring accurate high-resolution image based on its corresponding low-resolution version. However, most state-of-the-art SISR approaches attempt to achieve higher accuracy by pursuing deeper or more complicated models, which adversely increases computational cost. To achieve a good balance between restoration accuracy and computational speed, we make simple but effective modifications to the structure of residual blocks and skip-connections between stacked layers, and then propose a novel energy-aware training loss to adaptively adjust the restoration of high-frequency and low-frequency image regions. Extensive qualitative and quantitative evaluation results on benchmark datasets verify the effectiveness of the proposed techniques that they significantly improve SISR accuracy while causing no/ignorable extra computational loads.

Advances in low-frequency ultrasound combined with microbubbles in targeted tumor therapy.

The development of low-frequency ultrasound imaging technology and the improvement of ultrasound contrast agent production technology mean that they play an increasingly important role in tumor therapy. The interaction between ultrasound and microbubbles and their biological effects can transfer and release microbubbles carrying genes and drugs to target tissues, mediate the apoptosis of tumor cells, and block the embolization of tumor microvasculature. With the optimization of ultrasound parameters, the development of targeted microbubbles, and the emergence of various composite probes with both diagnostic and therapeutic functions, low-frequency ultrasound combined with microbubble contrast agents will bring new hope for clinical tumor treatment.

A New Pulse Coupled Neural Network (PCNN) for Brain Medical Image Fusion Empowered by Shuffled Frog Leaping Algorithm.

Recent research has reported the application of image fusion technologies in medical images in a wide range of aspects, such as in the diagnosis of brain diseases, the detection of glioma and the diagnosis of Alzheimer’s disease. In our study, a new fusion method based on the combination of the shuffled frog leaping algorithm (SFLA) and the pulse coupled neural network (PCNN) is proposed for the fusion of SPECT and CT images to improve the quality of fused brain images. First, the intensity-hue-saturation (IHS) of a SPECT and CT image are decomposed using a non-subsampled contourlet transform (NSCT) independently, where both low-frequency and high-frequency images, using NSCT, are obtained. We then used the combined SFLA and PCNN to fuse the high-frequency sub-band images and low-frequency images. The SFLA is considered to optimize the PCNN network parameters. Finally, the fused image was produced from the reversed NSCT and reversed IHS transforms. We evaluated our algorithms against standard deviation (SD), mean gradient (Ḡ), spatial frequency (SF) and information entropy (E) using three different sets of brain images. The experimental results demonstrated the superior performance of the proposed fusion method to enhance both precision and spatial resolution significantly.