Handwritten Signature Identification Based on Hybrid Features and Machine Learning Algorithms

ABSTRACTObjectiveIn spinal surgery, precise identification of high‐speed bur milling states is crucial for patient safety. This study investigates whether integrating tactile and auditory perception can enhance the accuracy of milling state detection in robot‐assisted cervical laminectomy.MethodsBased on the mathematical and physical model of vibration and sound in high‐speed bur milling bone, the feasibility of employing vibration and sound characteristics to identify the milling states of high‐speed bur is studied systematically. Cervical laminectomy was performed on the cervical spine of the sheep. During the signal acquisition process, acceleration sensors and microphones were installed to collect vibration and sound signals, respectively. Seven milling states were set up in the experiment: (1) Milling depths of cortical bone (CTB): 0.5, 1.0, and 1.5 mm; (2) Milling depths of milling of cancellous bone (CCB): 0.5, 1.0, and 1.5 mm; (3) Boundary conditions: high‐speed bur idling or complete penetration of bone (PT). The milling speed was set at 0.5 mm/s, the milling angle was 45°, and the bur diameter was 4 mm. The vibration or sound was extracted by Fast Fourier Transform (FFT) in the frequency domain of the first nine harmonics to generate the feature vector in 9 dimensions (9‐D) space. These vibration and sound features were combined to form an 18‐D multi‐perception spatial vector for subsequent analysis, including five machine learning algorithms: Support Vector Machine (SVM), K Nearest Neighbors (KNN), Naive Bayes (NB), Linear Discriminant Analysis (LDA), and Decision Tree (DT), and deep learning models: Long Short‐Term Memory networks (LSTM).ResultsBased on the 18‐D features of tactile and auditory multisensory fusion, the LSTM model is trained using 6600 sets of high‐speed bur milling data. In order to achieve the best performance, a layer‐by‐layer parameter optimization strategy was used to determine the optimal parameter configuration, and finally, a single‐layer LSTM with 12 memory units was constructed. In terms of accuracy and stability, the model is significantly superior to the machine learning algorithms (SVM, KNN, NB, LDA, and DT), and the accuracy of LSTM is 99.32% in the milling states identification of cervical lamina milling with high‐speed bur.ConclusionThrough theoretical analysis and experimental verification, the study built a multi‐perception fusion framework based on tactile and auditory perception and accurately identified the cervical vertebra milling states through the LSTM model, which can provide perception means for operational spinal surgery robots in the future.

A computer vision system for automatic recognition and classification of five varieties of plant leaves under controlled laboratory imaging conditions, comprising: 1–Cydonia oblonga (quince), 2–Eucalyptus camaldulensis dehn (river red gum), 3–Malus pumila (apple), 4–Pistacia atlantica (mt. Atlas mastic tree) and 5–Prunus armeniaca (apricot), is proposed. 516 tree leaves images were taken and 285 features computed from each object including shape features, color features, texture features based on the gray level co-occurrence matrix, texture descriptors based on histogram and moment invariants. Seven discriminant features were selected and input for classification purposes using three classifiers: hybrid artificial neural network–ant bee colony (ANN–ABC), hybrid artificial neural network–biogeography based optimization (ANN–BBO) and Fisher linear discriminant analysis (LDA). Mean correct classification rates (CCR), resulted in 94.04%, 89.23%, and 93.99%, for hybrid ANN–ABC; hybrid ANN–BBO; and LDA classifiers, respectively. Best classifier mean area under curve (AUC), mean sensitivity, and mean specificity, were computed for the five tree varieties under study, resulting in: 1–Cydonia oblonga (quince) 0.991 (ANN–ABC), 95.89% (ANN–ABC), 95.91% (ANN–ABC); 2–Eucalyptus camaldulensis dehn (river red gum) 1.00 (LDA), 100% (LDA), 100% (LDA); 3–Malus pumila (apple) 0.996 (LDA), 96.63% (LDA), 94.99% (LDA); 4–Pistacia atlantica (mt. Atlas mastic tree) 0.979 (LDA), 91.71% (LDA), 82.57% (LDA); and 5–Prunus armeniaca (apricot) 0.994 (LDA), 88.67% (LDA), 94.65% (LDA), respectively.

Machine learning (ML) models are cost-effective tools that can establish a prognosis by considering clinical, genomic, and histological data from breast cancer patients. In this study, we evaluate the performance of 12 ML models in predicting the overall survival of breast cancer patients from “The Cancer Genome Atlas-Breast Cancer” database at 60 months of follow-up. The models evaluated were Logistic Regression (LR), Ridge Classifier (RC), Least Absolute Shrinkage and Selection Operator (LASSO), K-Nearest Neighbors (KNN), Naive Bayes (NB), Linear Discriminant Analysis (LDA), Decision Tree (DT), Multilayer Perceptron (MP), Stochastic Gradient Descent (SGD), Support Vector Machine (SVM), Random Forest (RF), and XGBoost (XGB). The most accurate models were NB (86.76%) and SGD (85.29%). The models with the highest sensitivity were SGD (93.75%), MP (91.67%), and NB (89.89%), while the models with the highest specificity were KNN (95.0%), DT (90.0%), and LASSO (85.0%). The models with the highest area under the ROC curve were LDA (90.83%), LASSO (90.1%), and LR (89.79%). The most relevant variables were “Previous Diagnosis of Cancer,” “Presence of Tumor,” “Ancillary Therapy,” and “Histology.” Our findings demonstrate that AI tools for predicting patient outcomes are a viable option when expensive prognostic genomic tools are not available. According to our results, NB and LASSO show promising performance in predicting the prognosis of breast cancer patients. Further studies with data from developing countries are needed to improve the performance of these tools.

Biometrics is the study of unique characteristics present in the human body such as fingerprint, palm-print, retina, iris, footprint, etc. While other traits have been explored widely, only a few people have been considered the foot-palm region, despite having unique properties. Prior work has explored the foot shape features using length, width, major axis, minor axis, centroid, etc. but they are not reliable for personal verification due to similarity in the physical composition of two persons. It increases the demand for more unique features based on the footprint. Footprint texture features coming from creases of foot palm are unique and permanent like palmprint texture features. Hence the main objective of the paper is to investigate various kinds of texture feature techniques. These techniques will be further used in correct extraction of footprint features. After extraction of footprint features a detailed experimental analysis is performed to discover the uniqueness in foot texture. It is further utilized to test its viability as a human recognition trait. We describe a detailed feature extraction and classification technique applied to a collected footprint data-set. For feature extraction, we use three techniques: Gray Level Co-occurrence Matrix (GLCM), Histogram Oriented Gradient (HOG), and Local Binary Patterns (LBP). Feature classification is performed using four techniques: Linear Discriminant Analysis (LDA), Support Vector Machine (SVM), K-Nearest Neighbor (KNN), and Ensemble Subspace Discriminant (ESD). GLCM provides less accuracy, while HOG generates a big feature vector which takes more execution time. LBP provides a trade-off between the accuracy and the execution time. Detailed quantitative experiments show: GLCM with LDA provides an accuracy of $$88.5\%$$ , HOG with Fine-KNN achieves $$86.5\%$$ accuracy and LBP with LDA achieves the accuracy of $$97.9\%$$ .

This paper introduces a new face recognition method based on the gray-level co-occurrence matrix (GLCM). Both distributions of the intensities and information about relative position of neighbourhood pixels are carried by GLCM. Two methods have been used to extract feature vectors from the GLCM for face classification. The first, method extracts the well-known Haralick features to form the feature vector, where the second method directly uses GLCM by converting the matrix into a vector which can be used as a feature vector for the classification process. The results demonstrate that using the GLCM directly as the feature vector in the recognition process outperforms the feature vector containing the statistical Haralick features. Additionally, the proposed GLCM based face recognition system outperforms the well-known techniques such as principal component analysis and linear discriminant analysis.

Data classification in the system is widely used in various fields of information technology, data classification is obtained by analyzing objects based on the level of similarity and mapping the characteristics that are protected in the object based on the specified cluster. The use of classification in this study by use texture patterns contained in data by use K-Means Clustering algorithm and Naive Bayes Classifier. The K-Means Clustering algorithm is the process of grouping data based on the same characteristics, while Naive Bayes Classifier use probability and statistic calculations. This study aims to analyze data through texture patterns and the level of accuracy generated through K-Means Clustering algorithms and naïve bayes classifiers. Testing data on textures based on extract parameters with contrast and correlation classifications in GLCM (Gray-Level Co-Occurrence Matrix) methods and calculating the distance expressed in pixels with angle intervals of 0º, 45º, 90º, and 135º. Texture imagery is classified into three clusters by comparing accuracy values obtained. The accuracy value resulting from testing using the K-means Classifier algorithm is very good with an accuracy value of 100% and the accuracy value resulting from testing using the Naive Bayes Classifier method is very good with an accuracy value of 100%. Keywords: K-Means Clustering, Naive Bayes, and Classifier

The performance of seven classification methods, binary logistic (BLR), probit (PR) and cumulative probit (CPR) regression, linear (LDA) and quadratic (QDA) discriminant analysis, artificial neural networks (ANN), and naïve Bayes classification (NBC), is examined in skeletal sex estimation. These methods were tested using cranial and pelvic sexually dimorphic traits recorded on a modern documented collection, the Athens Collection. For their implementation, an R package has been written to perform cross-validated (CV) sex classification and give the discriminant function of each of the methods studied. A simple algorithm that combines two discriminant functions is also proposed. It was found that the differences in the classification performance between BLR, PR, CPR, LDA, QDA, ANN, and NBC are overall small. However, LDA is simpler and more flexible than CPR, QDA, and ANN and has a small but clear advantage over BLR, NBC, and PR. Consequently, LDA may be preferred in skeletal sex estimation. Finally, it is striking that the combination of pelvic and cranial traits via their discriminant functions, determined either by BLR or LDA, removes practically any population-specificity and yields much better predictions than the individual functions; in fact, the prediction accuracy increases above 97%.

Image analysis-based classification of pigmentation patterns in fish: A case study of pseudo-albinism in Senegalese sole

Application of Gabor wavelet and Locality Sensitive Discriminant Analysis for automated identification of breast cancer using digitized mammogram images

Computational approach for chronic wound tissue characterization

Aim) Pathological brain detection (PBD) systems aim to assist and even replace neuroradiologists to make decisions for patients. This review offers a comprehensive and quantitative comparison for PBD systems by artificial intelligence in magnetic resonance imaging (MRI) scanning. (Method) We first investigated four categories of brain diseases, including neoplastic disease, neurodegenerative disease, cerebrovascular disease, and inflammation. Next, we introduced important MRI techniques, such as the shimming, water and fat suppression, and three advanced imaging modalities (functional MRI, diffusion tensor imaging, and magnetic resonance spectroscopic imaging). Then, we discussed four image preprocessing techniques (image denoising, slice selection, brain extraction, spatial normalization, and intensity normalization), seven feature representation techniques (shape, moment, wavelet, statistics, entropy, gray level co-occurrence matrix, and Fourier transform), and two dimension reduction techniques (feature selection and feature extraction). Afterwards, we studied classification related methods: six learning models (decision tree, extreme learning machine, k-nearest neighbors, naive Bayes classifier, support vector machine, feed-forward neural network), five kernel functions (linear, homogeneous and inhomogeneous polynomial, radial basis function, and sigmoid), and three types of optimization methods (evolutionary algorithm, stochastic optimization, and swarm intelligence). (Results) We introduced three benchmark datasets and used K-fold stratified cross validation to avoid overfitting. We presented a detailed quantitative comparison among 44 state-of-theart PBD algorithms and discussed their advantages and limitations. (Discussions) Artificial intelligence is now making stride in the PBD field and enjoys a fair amount of success. In the future, semi-supervised learning and transfer learning techniques may be potential breakthroughs to develop PBD systems.

To develop an automatic tool on screening diabetic retinopathy (DR) from diabetic patients. We extracted textures from eye fundus images of each diabetes subject using grey level co-occurrence matrix method and trained a Bayesian model based on these textures. The receiver operating characteristic (ROC) curve was used to estimate the sensitivity and specificity of the Bayesian model. A total of 1000 eyes fundus images from diabetic patients in which 298 eyes were diagnosed as DR by two ophthalmologists. The Bayesian model was trained using four extracted textures including contrast, entropy, angular second moment and correlation using a training dataset. The Bayesian model achieved a sensitivity of 0.949 and a specificity of 0.928 in the validation dataset. The area under the ROC curve was 0.938, and the 10-fold cross validation method showed that the average accuracy rate is 93.5%. Textures extracted by grey level co-occurrence can be useful information for DR diagnosis, and a trained Bayesian model based on these textures can be an effective tool for DR screening among diabetic patients.

The analysis of digital pathology images, catalyzes research and automate diagnosis for improving clinical care. Regardless of the advances in high-resolution and speedy scanning devices, the clinical analysis of Whole Slide Image (WSI) remains mainly as the work of an image analyst. The complications when working on these images are primarily, its enormous image size, as well as the heterogeneous nature of tissue regions. These reasons may lead to inefficient computer analysis. We propose a comprehensive algorithm to analyse, these large images by using small tile image analysis. The small tile images can be learnt individually in detail, rather than processing the whole high resolution WSI. The analysis of these small tile images helps us configure, the localised pathology image characteristics. The work flow involves extracting Perceptual Features (PF), Fourier Local Binary Patterns (FLBP), Histogram, Gabor and Gray Level Co-occurrence Matrix (GLCM) features from the tiled WSI. Then, the multidimensional feature set is provided to the classifiers like K-Nearest Neighbour (KNN), Support Vector Machine (SVM), Naive Bayes (NB) and Logistic Regression (LR) for performance analysis. We evaluated our predictive models by automatically classifying image datasets (186 and 124) of brain tumour cases, using 10 fold cross validation. The model classified the images into Low Grade Glioma (LGG) and Glioblastoma Multiforme (GBM). It gives the highest accuracy of 93.5% for a feature combination of PF+FLBP+GLCM+GABOR with linear Support Vector Machine (linSVM) and Area under the Curve (AUC) of 0.9741 for the 186 image dataset. Also linSVM gives an accuracy of 96.8% with various feature combinations for the 124 image dataset. We did an exhaustive study on the performance of various sets of texture descriptors and classification models. This resulted in finding an optimised classification approach with a combined feature set that outperformed the original feature set classification. The proposed work is unique because it combines a wide range of feature extracting methods and an exhaustive list of classifiers showcasing the best method. The idea that, rather than selecting a region of interest from WSI for grade designation, all the regions in a slide are considered. This enables the involvement of each tile in making the final diagnostic decision, considering the heterogeneous nature of WSI. Our work can be extended for different tumour subtypes. Also, the experiment can be scaled with different sample sizes and finally the method works without a pathologist input.

A signature is one of the official biometric identification media that has special characteristics that are difficult to distinguish just by comparing with existing signatures. This causes irresponsible people to use signatures as objects of forgery. In addition, handwritten signature identification is a challenge because it has less information when compared to digital signatures. This research proposes a new approach to identify handwritten signature forgeries using a convolutional neural network (CNN) using mobile-based. Two model architectures are compared with the same testing scenario. The first architecture is a simplification of the Siamese convolutional network namely adjusted CNN. The second architecture is a standard CNN with 6 convolution layers which in this study is called general CNN. CNN is used for both feature extraction and classification processes. The dataset is built with a limited number of individuals consisting of genuine and forged signatures. Based on the experiments, the accuracy of each network reached 88.89% for adjusted CNN and 84.45% for general CNN. Adjusted CNN then embedded in a mobile-based application to identify the signature directly. The application that is built makes the process of identifying signatures more ease.

Accuracy of discriminant analysis methods to classify activities of subjects with Parkinson's Disease using wearable sensors