Two-dimensional Discrete Wavelet Transform Research Articles

Handwritten Character Recognition using Deep Learning Algorithm with Machine Learning Classifier

Handwritten character recognition is a problem that has been worked on for many mainstream languages. Handwritten letter recognition has been proven to achieve promising results. Several studies using deep learning models have been conducted to achieve better accuracies. In this paper, the authors conducted two experiments on the EMNIST Letters dataset: Wavemix-Lite and CoAtNet. The Wavemix-Lite model uses Two-Dimensional Discrete Wavelet Transform Level 1 to reduce the parameters and speed up the runtime. The CoAtNet is a combined model of CNN and Visual Transformer where the image is broken down into fixed-size patches. The feature extraction part of the model is used to embed the input image into a feature vector. From those two models, the authors hooked the value of the features of the Global Average Pool layer using EMNIST Letters data. The features hooked from the training results of the two models, such as SVM, Random Forest, and XGBoost models, were used to train the machine learning classifier. The experiments conducted by the authors show that the best machine-learning model is the Random Forest, with 96.03% accuracy using the Wavemix-Lite model and 97.90% accuracy using the CoAtNet model. These results showcased the benefit of using a machine learning model for classifying image features that are extracted using a deep learning model.

Multi-spatial scale and multi-frequency resolution Proper Orthogonal Decomposition for patterns of wall-streamwise pressure gradient fluctuation of impinging jets

In the coating industry, seemingly ordered millimeter-scale W or X-like patterns inevitably appear on the coating wiped by jets. Considering the dynamic control of coating thickness by the maximum wall-streamwise pressure gradient on the impinging wall, exploring its fluctuation characteristics may provide insights into W or X-like patterns formation and suppression. Large Eddy Simulation is conducted to simulate jets impinging on a dry wall. The computed pressure gradient is processed by a two-dimensional Discrete Wavelet Transform using Coif2 wavelet, helping determine the spatial scale dominated by its fluctuation energy and providing snapshots on specific spatial scales. Afterward, snapshots on selected spatial scales undergo processing using either multi-scale Proper Orthogonal Decomposition (mPOD) or spectral POD, depending on the presence of some distinguishable peaks in its spectrum. This process can extract modes with spectral purity on the corresponding spatial scale, providing an advantage in addressing problems strongly linked to spatial scales. Applying this methodology, the study unveiled the evolution of the pressure gradient modes with frequency on the energy-dominant spatial scale, ranging from 0.625 mm × 0.785 mm to 1.25 mm × 1.57 mm. This evolution may be the cause of W or X-like patterns and suggests introducing a guiding jet to mitigate them.

Harmonizing Cross View Image Transformation Through Local and Global Insights- A Review

Hyperspectral and multispectral information processing systems and technologies have demonstrated their value in enhancing agricultural productivity and practices by providing farmers and crop managers with valuable information on factors influencing crop condition and growth. These technologies play a crucial role in various agricultural applications, such as crop management, crop yield forecasting, crop disease detection, and monitoring soil, water, and land usage. To enhance the process of agriculture through effective crop management and assistance to farmer, using an advanced image transformation techniques the study delves into the exploration of techniques for harmonizing cross-view image transformation, with a focus on integrating both local and global insights. Further the study proposes the application of the 2D-DWT (Two-Dimensional Discrete Wavelet Transform) technique for image data preprocessing and the Extreme Learning Machine (ELM) algorithm for image classification in the context of hyperspectral images. Hyperspectral information sensing allows for the coverage of the electromagnetic spectrum in a single acquisition with several hundred spectral bands, resulting in a data cube containing significant spectral and spatial information ELM is particularly effective for classification problems because of its quick training time and capacity to operate with hidden nodes whose parameters are randomly assigned rather than modified iteratively. This comprehensive review aims to provide valuable insights and a critical analysis of the suggested method, shedding light on its potential contributions to the advancement of image transformation techniques for agricultural development.

Recognition of REEs anomalies using an image Fusion fractal-wavelet model in Tarom metallogenic zone, NW Iran

This study aims to detect REE geochemical anomalies in relationship to Iron-apatite ores utilizing an image Fusion based on Deep Learning (FDL). The geochemical data was modeled for elements related to Iron-apatite mineralization using multi b Spline B. The results were fusioned in applying the Deep learning method based on pre-trained networks. Wavelet-Number (WN) fractal model classified the best results based on the combination of a two-dimensional Discrete Wavelet Transformation (DWT) signal analysis and a Concentration-Area (C-A) fractal modeling. Sym8 carried the DWT as a selected wavelet pattern for REE based on Stream sediment samples collected from the Tarom region (NW Iran). In addition, the DWT was decomposed by wavelet coefficients at five levels. Furthermore, the DWT data were classified using a fractal-wavelet model to delineate REE anomalies from background levels in this region. Overlayed with the catchment basins model and weighted using the upstream and downstream parts. As a result, the prominent REE source anomalies are located in the southern parts of the study area. The results obtained by the proposed fractal-wavelet modeling are in connection with field check anomaly samples and the rock samples collected from the Iron-Apatite ore deposits.

Few-shot learning CNN optimized using combined 2D-DWT injection and evolutionary optimization algorithms for human face recognition

Convolutional Neural Network (CNN) has shown remarkable success in the area of machine vision. The purpose of this research is to enhance the classification for the few-shot learning datasets by developing a robust feature extraction system using an optimized CNN model. The aforementioned goal is attained in the following way by developing two classification models, (1) CNN optimized using Two-Dimensional Discrete Wavelet Transform (2D-DWT) injection using Principal Component Analysis (PCA), and Grey Wolf Optimizer (GWO), and (2) CNN optimized using 2D-DWT injection using PCA and Multi-Verse Optimizer (MVO) algorithm. This optimization process enhances the rate of face recognition for the small training dataset by extracting maximum features. Experiments on the AT&T (ORL), LFW, and Extended Yale-FACE-B databases show that the technique improves results significantly, with recognition rates increasing to 100% for training accuracy on all datasets and 100%, 98%, 97% on ORL, LFW, and Extended Yale-FACE-B datasets, respectively, for testing accuracy.

Hybrid Watermarking and Encryption Techniques for Securing Medical Images

Securing medical data while transmission on the network is required because it is sensitive and life-dependent data. Many methods are used for protection, such as Steganography, Digital Signature, Cryptography, and Watermarking. This paper introduces a novel robust algorithm that combines discrete wavelet transform (DWT), discrete cosine transform (DCT), and singular value decomposition (SVD) digital image-watermarking algorithms. The host image is decomposed using a two-dimensional DWT (2D-DWT) to approximate low-frequency sub-bands in the embedding process. Then the sub-band low-high (LH) is decomposed using 2D-DWT to four new sub-bands. The resulting sub-band low-high (LH1) is decomposed using 2D-DWT to four new sub-bands. Two frequency bands, high-high (HH2) and high-low (HL2), are transformed by DCT, and then the SVD is applied to the DCT coefficients. The strongest modified singular values (SVs) vary very little for most attacks, which is an important property of SVD watermarking. The two watermark images are encrypted using two layers of encryption, circular and chaotic encryption techniques, to increase security. The first encrypted watermark is embedded in the S component of the DCT components of the HL2 coefficients. The second encrypted watermark is embedded in the S component of the DCT components of the HH2 coefficients. The suggested technique has been tested against various attacks and proven to provide excellent stability and imperceptibility results.

Image Denoising using Wavelet Cycle Spinning and Non-local Means Filter

Removing as much noise as possible in an image while preserving its fine details is a complex and challenging task. We propose a wavelet-based and non-local means (NLM) denoising method to overcome the problem. Two well-known wavelets: dual-tree complex wavelet transform (DT-CWT) and discrete wavelet transform (DWT), have been used to change the noise image into several wavelet coefficients sequentially. NLM filtering and universal hard thresholding with cycle spinning have been used for thresholding on its approximation and detail coefficients, respectively. The inverse two-dimensional DWT was applied to the modified wavelet coefficients to obtain the denoised image. We conducted experiments with twelve test images on the set12 data set, adding the additive Gaussian white noise with variances of 10 to 90 in increments of 10. Three evaluation metrics, such as peak signal noise to rate (PSNR), structural similarity index metric (SSIM), and mean square error (MSE), have been used to evaluate the effectiveness of the proposed denoising method. From these measurement results, the proposed denoising method outperforms DT-CWT, DWT, and NLM almost in all noise levels except for the noise level of 10. At that noise level, the proposed denoising method is lower than NLM but better than DT-CWT and DWT.

Revisiting segmentation of lung tumors from CT images

Lung cancer is a leading cause of death throughout the world. Because the prompt diagnosis of tumors allows oncologists to discern their nature, type, and mode of treatment, tumor detection and segmentation from CT scan images is a crucial field of study. This paper investigates lung tumor segmentation via a two-dimensional Discrete Wavelet Transform (DWT) on the LOTUS dataset (31,247 training, and 4458 testing samples) and a Deeply Supervised MultiResUNet model. Coupling the DWT, which is used to achieve a more meticulous textural analysis while integrating information from neighboring CT slices, with the deep supervision of the model architecture results in an improved dice coefficient of 0.8472. A key characteristic of our approach is its avoidance of 3D kernels (despite being used for a 3D segmentation task), thereby making it quite lightweight.

Effective Classification of Synovial Sarcoma Cancer Using Structure Features and Support Vectors

In this research work, we proposed a medical image analysis framework with two separate releases whether or not Synovial Sarcoma (SS) is the cell structure for cancer. Within this framework the histopathology images are decomposed into a third-level sub-band using a two-dimensional Discrete Wavelet Transform. Subsequently, the structure features (SFs) such as Principal Components Analysis (PCA), Independent Components Analysis (ICA) and Linear Discriminant Analysis (LDA) were extracted from this sub-band image representation with the distribution of wavelet coefficients. These SFs are used as inputs of the Support Vector Machine (SVM) classifier. Also, classification of PCA + SVM, ICA + SVM, and LDA + SVM with Radial Basis Function (RBF) kernel the efficiency of the process is differentiated and compared with the best classification results. Furthermore, data collected on the internet from various histopathological centres via the Internet of Things (IoT) are stored and shared on blockchain technology across a wide range of image distribution across secure data IoT devices. Due to this, the minimum and maximum values of the kernel parameter are adjusted and updated periodically for the purpose of industrial application in device calibration. Consequently, these resolutions are presented with an excellent example of a technique for training and testing the cancer cell structure prognosis methods in spindle shaped cell (SSC) histopathological imaging databases. The performance characteristics of cross-validation are evaluated with the help of the receiver operating characteristics (ROC) curve, and significant differences in classification performance between the techniques are analyzed. The combination of LDA + SVM technique has been proven to be essential for intelligent SS cancer detection in the future, and it offers excellent classification accuracy, sensitivity, specificity.

Design of self-starting multivariate control chart for monitoring patients suspected to bone marrow metastasis

Nowadays, statistical process control methods are widely used in healthcare, especially for cancer patients, to help doctors in interpreting and diagnosing medical data. Since adequate data from patients with bone marrow metastasis are often unavailable, the researchers use a self-starting control chart. In this regard, using the self-starting multivariate control chart, we monitor the status of people suspected of bone marrow metastasis in the pelvic region. For this, using a two-dimensional discrete wavelet transformation, we extracted some features from the ADC and T1 magnetic resonance images of ten bone marrow metastasis samples. Out of these features, we selected six ones for final analysis. Then, using the self-starting SSMEWMA and SST2, we performed the simulation studies, as well as a numerical example, on the extracted features and evaluated the performance of the proposed methods in terms of average run length measure. The simulation results verified the appropriate performance of the proposed methods in diagnosing patients with bone marrow metastasis than non-metastasis ones.

Reversible Video Steganography Using Quick Response Codes and Modified ElGamal Cryptosystem

The rapid transmission of multimedia information has been achieved mainly by recent advancements in the Internet's speed and information technology. In spite of this, advancements in technology have resulted in breaches of privacy and data security. When it comes to protecting private information in today's Internet era, digital steganography is vital. Many academics are interested in digital video because it has a great capability for concealing important data. There have been a vast number of video steganography solutions developed lately to guard against the theft of confidential data. The visual imperceptibility, robustness, and embedding capacity of these approaches are all challenges that must be addressed. In this paper, a novel solution to reversible video steganography based on Discrete Wavelet Transform (DWT) and Quick Response (QR) codes is proposed to address these concerns. In order to increase the security level of the suggested method, an enhanced ElGamal cryptosystem has also been proposed. Prior to the embedding stage, the suggested method uses the modified ElGamal algorithm to encrypt secret QR codes. Concurrently, it applies two-dimensional DWT on the Y-component of each video frame resulting in Approximation (LL), Horizontal (LH), Vertical (HL), and Diagonal (HH) sub-bands. Then, the encrypted Low (L), Medium (M), Quantile (Q), and High (H) QR codes are embedded into the HL sub-band, HH sub-band, U-component, and V-component of video frames, respectively, using the Least Significant Bit (LSB) technique. As a consequence of extensive testing of the approach, it was shown to be very secure and highly invisible, as well as highly resistant to attacks from Salt & Pepper, Gaussian, Poisson, and Speckle noises, which has an average Structural Similarity Index (SSIM) of more than 0.91. Aside from visual imperceptibility, the suggested method exceeds current methods in terms of Peak Signal-to-Noise Ratio (PSNR) average of 52.143 dB, and embedding capacity 1 bpp.

Detection of geochemical anomalies using a fractal-wavelet model in Ipack area, Central Iran

The purpose of this study is to detect geochemical anomalies utilizing a Wavelet-Number (W-N) fractal model based on combination of a two-dimensional Discrete Wavelet Transformation (DWT) signal analysis and a Concentration-Area (C-A) fractal modeling. The DWT was carried by Sym8 as a selected wavelet pattern (by energy matching method) for Zn and Pb based on rock samples collected from the Ipack area, Central Iran. In addition, the DWT was decomposed by continuous wavelet coefficients at five levels and a final image was synthesized. Furthermore, the DWT data were classified using a fractal-wavelet model for delineation of both Zn and Pb anomalies from background levels in this area. Moreover, a Spectrum-Area (S-A) fractal modeling and classical statistics analysis were carried out and their results were compared with the wavelet-fractal modeling. As a result, the main Zn and Pb anomalies are located in the central and southern parts of the Ipack area. The results obtained by the proposed fractal-wavelet modeling are in connection with drilling for Zn and Pb exploration.

A Novel Image Encryption Scheme Based on 2D Fractional Chaotic Map, DWT and 4D Hyper-chaos

In this paper, a novel image encryption scheme based on a fractional-order Henon chaotic map, a two-dimensional (2D) Discrete Wavelet Transform (DWT) and a four-dimensional (4D) hyperchaotic system is proposed. Firstly, the original image is transformed and scrambled by the 2D DWT, and then the image is shuffled with the fractional-order Henon chaotic time series. Finally, the shuffled image is diffused and encrypted by the 4D hyperchaos system. Through the application of DWT and high-low dimensional chaotic systems, the encryption effect of this algorithm is better than those done by single or ordinary chaotic encryption algorithm, and it has a larger key space and higher security. The experimental tests show that the system has good statistical characteristics, such as histogram analysis, correlation coefficient analysis, key space and key sensitivity, information entropy analysis and so on. The encryption algorithm also passes the relevant security attack tests with good security.

Correction of amplitude scintillation effect in fully polarimetric SAR coherency matrix data

In this paper, we investigate the correction of ionospheric amplitude scintillations appearing as image artifacts in low frequency synthetic aperture radar (SAR) data. A method has been proposed to separate amplitude and phase components of fully polarimetric SAR (POLSAR) coherency matrix [T], and then implement a combination of multi-level, two-dimensional Discrete Wavelet Transform and Fast Fourier Transform (DWT-FFT) to correct the amplitude scintillation-induced stripes in [T] matrix. In the methodology, [T] is decomposed into component matrices of amplitude and relative phases. To the amplitude [T] matrix, a combination of 2-D DWT is performed, using the Haar wavelet, followed by a non-linear 2-D FFT correction strategy to remove amplitude scintillation stripes. Using DWT, each amplitude [T] matrix element is decomposed into low- and high-frequency components known as approximate and detailed coefficients respectively. On applying a 2-D FFT on the detailed coefficients, amplitude scintillation-induced stripes are easily identified in the frequency spectrum and then removed using 2-D FFT correction strategy of averaging-post-thresholding. The proposed method is tested on the [T] matrix of several POLSAR datasets acquired from Advanced Land Observation Satellite/Phased Array type L-band synthetic aperture radar (ALOS/PALSAR) and Advanced Land Observation Satellite-2/Phased Array type L-band synthetic aperture radar-2 (ALOS-2/PALSAR-2) satellites. The performance of the correction technique is analysed both, qualitatively and quantitatively. While improvements in the post-correction datasets are visually observed, a supervised Wishart classification also performed on the [T] matrices of undisturbed (reference), disturbed and corrected data indicates significant increase in overall accuracy (OA) and kappa coefficient (k̂) parameters. The classification accuracies of the disturbed data (OA = 69.31%, k̂=0.54) improved to OA = 76.49%, k̂=0.68 after applying the proposed correction approach, which is comparable to the reference data values (OA = 81.57%, k̂=0.71). The proposed method performs significantly better than the existing technique of a simple 2-D FFT approach (OA = 73.49%, k̂=0.62). The contribution of dominant scattering powers, obtained from the seven-component scattering power decomposition of POLSAR coherency matrix (7SD) model, also demonstrates post-correction improvements. The dominant scattering power for targets increased by ~3–9% using the proposed approach, shows an improvement of ~3% over the existing 2-D FFT technique.

Spectral DWT Multilevel Decomposition with Spatial Filtering Enhancement Preprocessing-Based Approaches for Hyperspectral Imagery Classification

In this paper, spectral–spatial preprocessing using discrete wavelet transform (DWT) multilevel decomposition and spatial filtering is proposed for improving the accuracy of hyperspectral imagery classification. Specifically, spectral DWT multilevel decomposition (SDWT) is performed on the hyperspectral image to separate the approximation coefficients from the detail coefficients. For each level of decomposition, only the detail coefficients are spatially filtered instead of being discarded, as is often adopted by the wavelet-based approaches. Thus, three different spatial filters are explored, including two-dimensional DWT (2D-DWT), adaptive Wiener filter (AWF), and two-dimensional discrete cosine transform (2D-DCT). After the enhancement of the spectral information by performing the spatial filter on the detail coefficients, DWT reconstruction is carried out on both the approximation and the filtered detail coefficients. The final preprocessed image is fed into a linear support vector machine (SVM) classifier. Evaluation results on three widely used real hyperspectral datasets show that the proposed framework using spectral DWT multilevel decomposition with 2D-DCT filter (SDWT-2DCT_SVM) exhibits a significant performance and outperforms many state-of-the-art methods in terms of classification accuracy, even under the constraint of small training sample size, and execution time.

Evaluation of surface texture of dried Hami Jujube using optimized support vector machine based on visual features fusion.

The surface texture of dried jujube fruits is a significant quality grading criterion. This paper introduced a novel visual feature fusion based on connected region density, texture features, and color features. The single-scale Two-Dimensional Discrete Wavelet Transform was used to perform single-scale decomposition and reconstruction of dried Hami jujube image before visual features extraction. The connected region density was extracted by the two different algorithms, whereas the texture features were extracted by Gray Level Co-occurrence Matrix and the color features were extracted by image processing algorithms. Based on selected features which obtained by correlation analysis of visual features, the accuracy rate of the optimized Support Vector Machine classification model was 96.67%. In comparing with Extreme Learning Machine classification model and other fusion methods, the optimized Support Vector Machine based on selected visual features fusion was better.

Face Recognition with Symmetrical Face Training Samples Based on Local Binary Patterns and the Gabor Filter

In the practical reality of face recognition applications, the human face can have only a limited number of training images. However, it is known that, in general, increasing the number of training images also increases the performance of face recognition systems. In this case, a new set of training samples can be generated from the original samples, using the symmetry property of the face. Although many face recognition methods have been proposed in the literature, a robust face recognition system is still a challenging task. In this paper, recognition performance was improved by using the property of face symmetry. Moreover, the effects of illumination and pose variations were reduced. A Two-Dimensional Discrete Wavelet Transform, based on the Local Binary Pattern, which is a new approach for face recognition using symmetry, has been presented. The method has three main stages, preprocessing, feature extraction, and classification. A Two-Dimensional Discrete Wavelet Transform with Single-Level and Gaussian Low-Pass Filter were used, separately, for preprocessing. The Local Binary Pattern, Gray Level Co-Occurrence Matrix, and the Gabor filter were used for feature extraction, and the Euclidean Distance was used for classification. The proposed method was implemented and evaluated using the Olivetti Research Laboratory (ORL) and Yale datasets. This study also examined the importance of the preprocessing stage in a face recognition system. The experimental results showed that the proposed method had a recognition accuracy of 100%, for both the ORL and Yale datasets, and these recognition rates were higher than the methods in the literature.

An Illumination Pre-processing Method Using the Enhanced Energy of Discrete Wavelet Transform for Face Recognition

ABSTRACTAutomatic face recognition is useful in a wide range of applications. The accuracy of a face recognition system is adversely affected due to illumination variations. This study presents an illumination pre-processing method, “Enhanced Energy Discrete Wavelet Transform”, to increase recognition accuracy of front view faces under varying illuminations. This method is implemented as follows. A two-dimensional Discrete Wavelet Transform (2D DWT) decomposes the original image into four frequency sub-bands, namely low-low, low-high, high-low and high-high. The energy of each sub-band is computed. A weighting scheme is employed to increase the energy. The weighting scheme computes weight of sub-band based on its energy and threshold. The new sub-bands are obtained by multiplying sub-band, weight and weighting factor. Then, the new four sub-bands are added to create enhanced energy DWT image. The face recognition is carried out using Gabor magnitude features. The experiments are conducted in facial images of Yale, Pose Illumination Expression (PIE) and Extended Yale B. The results proves that low-low sub-band has maximum weight. The performance analysis is carried out using different threshold and weighting factor values. The weighting factor increases energy of low-low sub-band. The analysis of energy enhancement shows increase in energy of 20.96% and 31.88% in CMU PIE and Extended Yale B, respectively. The increase of energy improves the brightness of image. The recognition accuracy in CMU PIE and Extended Yale B is 99.02% and 99.82%, respectively. The comparative analysis with the state-of-the-art methods proves that the proposed method is an effective illumination pre-processing method.

Block Energy Based Visual Features Using Histogram Of Oriented Gradient For Bimodal Hindi Speech Recognition

In this paper, new set of visual speech feature based on Histogram of Oriented Gradient (HOG) is proposed to improve the robustness of bimodal Hindi speech recognition. For extracting the visual features, energy per block using HOG is calculated by finding the gradient magnitude of the pixel in the cell for both x and y direction form the region of interest (ROI). The advantage of proposed scheme is that it has reduced the dimensionality of visual features vectors which can retain the full information of the lip region. For comparative study, four sets of visual feature; Set A (Two-Dimensional Discrete Cosine Transform feature (2D-DCT)), Set B (Two-Dimensional Discrete Wavelet Transform followed by DCT (2D-DWT-DCT)), Set C (Static-HOG) and Set D (Dynamic-HOG) are extracted from AMUAV corpus. Standard Mel Frequency Cepstral coefficients (MFCC) followed by static and dynamic (MFCC) features were used as baseline features. The maximum improvement in WRA (%) of 12.73% is reported over baseline features using proposed features sets.

Design Of Enhanced SQRT Carry Select Adder for VLSI Implementation Of 2d- Discrete Wavelet Transform

In this paper, two-dimensional Discrete Wavelet Transform (2-D DWT) based image compression is presented with the help of developed Enhanced square root carry select adder (ESCSLA). The Enhanced Carry select adder is replacing the RCA and BEC unit to D-latch circuit. It provides better results than compared to the conventional adders. DWT and ESCSLA circuit is designed through Very Large Scale Integration (VLSI) System design environment. Low power consumption, less area and high speed are the main concerns in VLSI System design. Hence, our aim is to reduce the hardware complexity of ESCSLA and improve the performance of DWT in terms VLSI concerns. In our work, three levels of decomposition are made for image compression. With the help of pixel values of image and Enhanced square root carry select adder circuit, 2-D DWT image compression technique is implemented through Verilog HDL design. The performance of ESCSLA circuit is better than conventional Binary to Excess1 (BEC) based Square Root Carry Select Adder (SQRT CSLA) in terms of silicon area and power consumption. Hence, the performance of DWT also increased when incorporating EHRCA circuit into addition process of DWT computation.