DWT Features Research Articles

DWT and LBP hybrid feature based deep learning technique for image splicing forgery detection

DWT and LBP hybrid feature based deep learning technique for image splicing forgery detection

A New Facial Expression Recognition Algorithm Based on DWT Feature Extraction and Selection

A New Facial Expression Recognition Algorithm Based on DWT Feature Extraction and Selection

An automated crack detection method for underwater structures based on multilevel DWT and LPQ feature generation

An automated crack detection method for underwater structures based on multilevel DWT and LPQ feature generation

A Real-Time Deep Machine Learning Approach for Sudden Tool Failure Prediction and Prevention in Machining Processes.

Tool Condition Monitoring systems are essential to achieve the desired industrial competitive advantage in terms of reducing costs, increasing productivity, improving quality, and preventing machined part damage. A sudden tool failure is analytically unpredictable due to the high dynamics of the machining process in the industrial environment. Therefore, a system for detecting and preventing sudden tool failures was developed for real-time implementation. A discrete wavelet transform lifting scheme (DWT) was developed to extract a time-frequency representation of the AErms signals. A long short-term memory (LSTM) autoencoder was developed to compress and reconstruct the DWT features. The variations between the reconstructed and the original DWT representations due to the induced acoustic emissions (AE) waves during unstable crack propagation were used as a prefailure indicator. Based on the statistics of the LSTM autoencoder training process, a threshold was defined to detect tool prefailure regardless of the cutting conditions. Experimental validation results demonstrated the ability of the developed approach to accurately predict sudden tool failures before they occur and allow enough time to take corrective action to protect the machined part. The developed approach overcomes the limitations of the prefailure detection approach available in the literature in terms of defining a threshold function and sensitivity to chip adhesion-separation phenomenon during the machining of hard-to-cut materials.

ResNet18 Supported Inspection of Tuberculosis in Chest Radiographs With Integrated Deep, LBP, and DWT Features

ResNet18 Supported Inspection of Tuberculosis in Chest Radiographs With Integrated Deep, LBP, and DWT Features

Disease detection of apple leaf with combination of color segmentation and modified DWT

In this paper, we proposed a machine learning and computer vision-based automated apple disease detection and recognition system based on leaf symptoms. The proposed method is composed of three parts: diseased region segmentation, feature extraction, and classification. We have segmented the infected portion of the leaf using L*a*b* space-based color segmentation method. Here, average color markers in a*b* space and the nearest neighbor method have been used for classifying each pixel into either healthy, infected, or background regions. We have extracted two types of features: one is the proposed DWT feature and another is L*a*b* space-based color histogram features. Horizontal feature fusion is performed to create the final feature vector. The feature vectors have been classified using several classifiers keeping Random Forrest as the base classifier. In this paper, the experiment is made on Plant Village dataset, where image of Apple Scab, Black Rot, and Cedar Apple Rust disease are taken for both training and testing our model. The fusion of proposed DWT and color histogram features is a novel approach in detecting and recognizing apple leaf disease, which got an accuracy of 98.63%.

AFD filter and E2N2 classifier for improving visualization of crop image and crop classification in remote sensing image

ABSTRACT Due to increasing global population, the need of focusing on land cover and land uses by human beings is felt. Nowadays, most of the researchers and scientists are found showing much attention in agricultural landscapes because the effectual land-cover (LC) mapping provides the correct information regarding the land uses and land cover. Crop classification is one of the essential issues in the agriculture field because it affects the government and private agencies economically. The number of methodologies exist to classify the crop images from the remotely sensed image, but the traditional methodologies do not perform well due to some factors, like blurring effect, noise, etc. We have proposed an Adaptive Fuzzy-Centred Denoising Filter (AFDF) and Extended Elman Neural Network (E2N2) to enhance visualization of crop image to crop classification in order to overcome traditional techniques RSI. The remote-sensing crop images have been considered in the proposed work, and the AFDF preprocesses the image to eradicate the noise. After that, the Jaccard Coefficient-based Shared Nearest Neighbour (JCSNN) segments the object as preprocessed images. The LTrP, DWT, GLCM, and shape features are extracted as the optimal features chosen by Enhanced Elephant Herding Optimization (E2HO). Lastly, the crop is classified using the E2N2 classifier. The high-resolution images were utilized for the experiments to validate the proposed work. At last, various techniques have been compared against the proposed system, and outcomes have shown significant enhancement in accuracy and reliability. The proposed system obtains 88.01% F-measure, 95.2% accuracy with lesser FPR values that are greater when analogized to all prevalent methods. The advantage of using proposed effective protocols, like Adaptive Fuzzy-centred Denoising Filter and also E2N2 classifier, boost-up the crop images’ visualization and also crop categorization in remote-sensing images.

Optimized cancer detection on various magnified histopathological colon images based on DWT features and FCM clustering

Optimized cancer detection on various magnified histopathological colon images based on DWT features and FCM clustering

Prediction and Analysis of Muscular Paralysis Disease using DWT and Hybrid Features

Genetic advancements have shown that ALS is not a single entity but consists of a collection of syndromes in which the motor neurons degenerate. Together with these multiple genetic etiologies, there is a broad variability in the disease’s clinical manifestations in terms of the age of symptom onset, site of onset, rate and pattern of progression, and cognitive involvement. In this paper, prediction of human paralysis is done based on extraction of features from the ALS dataset samples. The classification has been carried by two different Machine Learning based algorithms i.e., Gradient Boosting (GB), and neural Network (NN). The standard data set such as ALS has been used for this purpose. The classifier model has used 80% data as a training set and the remaining 20% of data as the test set. The result shows that GB and NN perform better with an accuracy of 98%. Based on the desired accuracy, this classification model serves better compared with existing models.

Robust Classification Technique for Hyperspectral Images Based on 3D-Discrete Wavelet Transform

Hyperspectral image classification is an emerging and interesting research area that has attracted several researchers to contribute to this field. Hyperspectral images have multiple narrow bands for a single image that enable the development of algorithms to extract diverse features. Three-dimensional discrete wavelet transform (3D-DWT) has the advantage of extracting the spatial and spectral information simultaneously. Decomposing an image into a set of spatial–spectral components is an important characteristic of 3D-DWT. It has motivated us to perform the proposed research work. The novelty of this work is to bring out the features of 3D-DWT applicable to hyperspectral images classification using Haar, Fejér-Korovkin and Coiflet filters. Three-dimensional-DWT is implemented with the help of three stages of 1D-DWT. The first two stages of 3D-DWT are extracting spatial resolution, and the third stage is extracting the spectral content. In this work, the 3D-DWT features are extracted and fed to the following classifiers (i) random forest (ii) K-nearest neighbor (KNN) and (iii) support vector machine (SVM). Exploiting both spectral and spatial features help the classifiers to provide a better classification accuracy. A comparison of results was performed with the same classifiers without DWT features. The experiments were performed using Salinas Scene and Indian Pines hyperspectral datasets. From the experiments, it has been observed that the SVM with 3D-DWT features performs better in terms of the performance metrics such as overall accuracy, average accuracy and kappa coefficient. It has shown significant improvement compared to the state of art techniques. The overall accuracy of 3D-DWT+SVM is 88.3%, which is 14.5% larger than that of traditional SVM (77.1%) for the Indian Pines dataset. The classification map of 3D-DWT + SVM is more closely related to the ground truth map.

Pashto Characters Recognition Using Multi-Class Enabled Support Vector Machine

During the last two decades significant work has been reported in the field of cursive language’s recognition especially, in the Arabic, the Urdu and the Persian languages. The unavailability of such work in the Pashto language is because of: the absence of a standard database and of significant research work that ultimately acts as a big barrier for the research community. The slight change in the Pashto characters’ shape is an additional challenge for researchers. This paper presents an efficient OCR system for the handwritten Pashto characters based on multi-class enabled support vector machine using manifold feature extraction techniques. These feature extraction techniques include, tools such as zoning feature extractor, discrete cosine transform, discrete wavelet transform, and Gabor filters and histogram of oriented gradients. A hybrid feature map is developed by combining the manifold feature maps. This research work is performed by developing a medium-sized dataset of handwritten Pashto characters that encapsulate 200 handwritten samples for each 44 characters in the Pashto language. Recognition results are generated for the proposed model based on a manifold and hybrid feature map. An overall accuracy rates of 63.30%, 65.13%, 68.55%, 68.28%, 67.02% and 83% are generated based on a zoning technique, HoGs, Gabor filter, DCT, DWT and hybrid feature maps respectively. Applicability of the proposed model is also tested by comparing its results with a convolution neural network model. The convolution neural network-based model generated an accuracy rate of 81.02% smaller than the multi-class support vector machine. The highest accuracy rate of 83% for the multi-class SVM model based on a hybrid feature map reflects the applicability of the proposed model.

Low dimensional DCT and DWT feature based model for detection of image splicing and copy-move forgery

Digital images are being used as a prominent carrier of visual information in this age of digitization. Images become more and more omnipresent in everyday life. The images can be easily manipulated due to the accessibility of many internet tools and advanced software. Previously many techniques have been developed to authenticate the images. But all the previous techniques have high dimension of feature vectors. Here, a low dimensional DCT and DWT based features have been introduced to authenticate the images. In this work, we are dealing with both the passive forgery (splicing and copy-move) simultaneously. Features are extracted through image statistics and pixel correlation from DCT and DWT domain. Ensemble classifier has been selected for training and testing. The classifier classifies whether the given images are forged or authentic. Further, it also classifies the forgery in spliced or copy-move. If there is copy-move, the proposed work also perform the region detection using a novel key-point based method. The proposed model gives good detection accuracy and high generalization capability which is independent of image formats. Experimental results demonstrate the performance of proposed work against different post-processing operations like scaling, rotation, and Gaussian noise. Also, the comparative results against different existing methods show the effectiveness of the proposed model.

Optimized Multi-Model Biometric Based Human Authentication using Deep Neural Network

Biometrics provides greater security and usability than conventional personal authentication methods. Fingerprints, facial identification systems and voice recognition systems are the features that biometric systems can use. To improve biometric authentication, the proposed method considered that the input image is iris and fingerprint; at first, pre-processing is performed through histogram equalization for all image inputs to enhance the image quality. Then the extraction process of the feature will be performed. The suggested method uses modified Local Binary Pattern (MLBP), GLCM with orientation transformation, and DWT features next to the extracted features to be combined for feature extraction. Then the optimum function is found with the Rider Optimization Algorithm (ROA) for all MLBP, GLCM and DWT. Eventually, the approach suggested is accepted. Deep Neural Network (DNN) performs the proposed authentication process. A DNN is a multilayered artificial neural network between the layers of input and output. The DNN finds the right mathematical manipulation to turn the input into the output, whether it is an acknowledged image or not. Suggested process quality is measured in terms of reliability recognition. In the MATLAB platform, the suggested approach is implemented.

Classification of Tumors in Brain MRI Images With Hybrid of Global and Local DWT Features using Decision Tree

Automated brain tumor identification and classification is still an open problem for research in the medical image processing domain. Brain tumor is a bunch of unwanted cells that develop in the brain. This growth of a tumor takes up space within skull and affects the normal functioning of brain. Automated segmentation and detection of brain tumors are important in MRI scan analysis as it provides information about neural architecture of brain and also about abnormal tissues that are extremely necessary to identify appropriate surgical plan. Automating this process is a challenging task as tumor tissues show high diversity in appearance with different patients and also in many cases they tend to appear very similar to the normal tissues. Effective extraction of features that represent the tumor in brain image is the key for better classification. In this paper, we propose a hybrid feature extraction process. In this process, we combine the local and global features of the brain MRI using first by Discrete Wavelet Transformation and then using texture based statistical features by computing Gray Level Co-occurrence Matrix. The extracted combined features are used to construct decision tree for classification of brain tumors in to benign or malignant class.

An Efficient and Automatic ECG Arrhythmia Diagnosis System using DWT and HOS Features and Entropy- Based Feature Selection Procedure.

Primary recognition of heart diseases by exploiting computer aided diagnosis (CAD) machines, decreases the vast rate of fatality among cardiac patients. Recognition of heart abnormalities is a staggering task because the low changes in ECG signals may not be exactly specified with eyesight. In this paper, an efficient approach for ECG arrhythmia diagnosis is proposed based on a combination of discrete wavelet transform and higher order statistics feature extraction and entropy based feature selection methods. Using the neural network and support vector machine, five classes of heartbeat categories are classified. Applying the neural network and support vector machine method, our proposed system is able to classify the arrhythmia classes with high accuracy (99.83%) and (99.03%), respectively. The advantage of the presented procedure has been experimentally demonstrated compared to the other recently presented methods in terms of accuracy.

Normal and Abnormal Detection for Knee Osteoarthritis using Machine Learning Techniques

Osteoarthritis is the most broadly recognized disease in the knee joint that affects the cartilage, especially among the old age or overweight people. In the normal knee joint, the smooth and thin layer called cartilage covers the joint space of the bone and makes the joint smooth and prevents them from rubbing against one another, but can break, when the cartilage gets ruptured due to which bones start rubbing with one another, and this may cause severe pain, swelling and stiffness in the knee joint. The evaluation for osteoarthritis detection includes a clinical examination, and different medical imaging techniques are X-RAY images and MRI scans. There is developing method required for classification frameworks that can precisely distinguish and identify knee OA from plain radiographs. In this method we have examining the strategy of computer aided diagnosis for early identification of knee OA. Based on the procedure of x rays through computer image processing, segmentation, feature extraction and investigation by means of building a classifier, a viable computer aided detection method for knee was made to help specialists in their precise, convenient and identification of potential risk of OA. For this method a total of 126 knee x ray image were collected for assessing the knee OA. In this paper, we tried to diagnose about the normal or abnormal detection of cartilage depreciation. The HOG and DWT features are extracted from X-ray images of the knee joints. The extracted features are classified with two different machine learning classifiers, namely the SVM and ANN Patternet classifiers, and the results are demonstrated. The SVM classification is good when compared with ANN and provides a satisfactory accuracy rate of 85.33%. At last the classifier was superior both in time effectiveness and classification execution to the regularly utilized classifiers based on iterative learning. In this way it was suitable to utilize as a computer aided tool for the diagnosis of OA.

The Analysis of Shape-based, DWT and Zernike Moments Feature Extraction Techniques for Fasterner Recognition Using 10-Fold Cross Validation Multilayer Perceptrons

This paper presents an analysis of three feature extraction techniques which are the shape-based, Zernike moments and Discrete Wavelet Transform for fastener recognition. RGB colour features are also added to these major feature extractors to enhance the classification result. The classifier used in this experiment is back propagation neural network and the result in general is strengthen using ten-fold cross validation. The result is measured using percentage accuracy and Kappa statistics. The overall results showed that the best feature extraction techniques are Zernike moment group 3 and DWT both with added colour features.

Automatic Inshore Ship Detection in Satellite Imageries Using DWT and SIFT Features

Automatic Inshore Ship Detection in Satellite Imageries Using DWT and SIFT Features

Perceptual Image Hashing with Weighted DWT Features for Reduced-Reference Image Quality Assessment

Perceptual Image Hashing with Weighted DWT Features for Reduced-Reference Image Quality Assessment

Brain Tumor Classification Techniques using MRI: A Study

Medical sciences have a major role in every person's life. From children to old age people, all need treatment for various diseases. Though it has a tremendous impact, it still lacks cure to many diseases, one of which is brain tumor. Brain tumor can be dangerous if it is a malignant tumor or secondary tumor and needed proper evaluation for their treatment. Clinically, the efficiency and accuracy are important in such treatment. For this purpose, automated brain tumor detection can be use, where we can detect and classify the tumor without human intervention. Brain tumor detection and classification is an important research area that still lacks efficiency because of variations of tissue characteristics inside the brain like shape, size, location, etc. Many people are living with a brain tumor which is expected to be cured soon through diagnosis. Taking into account the importance of detection of brain tumor, this paper is focused on the classification of brain MRI images into tumorous or non-tumorous images. It also provides a comparison of different existing classification techniques using GLCM features and hybridization of GLCM and DWT features. Performance is evaluated using metrics like accuracy, precision, recall and Fmeasure. From our observation, we found that using hybridized GLCM and DWT feature, result is better than that of GLCM feature for most methodologies.