This paper presents a new automatic image enhancement method by modifying the gamma value of its individual pixels. Most of existing gamma correction methods apply a uniform gamma value across the image. Considering the fact that gamma variation for a single image is actually nonlinear, the proposed method locally estimates the gamma values in an image using support vector machine. First, a database of training images are constructed from various standard images under different gamma conditions. Then by windowing each of the training images, a number of features that characterize images content are computed from its pixel intensity histogram, gray level co-occurrence matrix, and discrete cosine transform domain. To improve the gamma values of an image the aforementioned features are initially computed in sliding windows, then SVM is employed to estimate the gamma value in each window. In this study, it is shown that the proposed method has performed well in improving the quality of images. Subjective and objective image quality assessments used in this study attest superiority of the proposed method compared to the existing methods in image quality enhancement using image gamma value.

Todays, Intrusion Detection Systems (IDS) are considered as key components of security networks. However, high false positive and false negative rates are the important problems of these systems. On the other hand, many of the existing solutions in the articles are restricted to class datasets due to the use of a specific technique, but in real applications they may have multi-variant datasets. With the impetus of the facts, this paper presents a new anomaly based intrusion detection system using J48 Decision Tree, Support Vector classifier (SVC) and k-means clustering algorithm in order to reduce false alarm rates and enhance the system performance. J48 decision tree algorithm is used to select the best features and optimize the dataset. Also, an SVM classifier and a modified k-means clustering algorithm are used to build a profile of normal and anomalous behaviors of dataset. Simulation results on benchmark NSL-KDD, CICIDS2017 and synthetic datasets confirm that the proposed method has significant performance in comparison with previous approaches.

Nowadays, IDS is an essential technology for defense in depth. Researchers have interested on IDS using data mining and artificial intelligence (AI) techniques as an artful. IDSs can monitor system behavior and network traffic until detect intrusive action. One of the IDS models is anomaly based IDS which trained to distinguish between normal and abnormal traffic. This paper has proposed an anomaly based IDS using GA for optimizing feature vectors and SVM as a classifier. SVM has used as a supervised learning machine that analyses data and recognize patterns, used for classification and regression analysis. After optimization best features for SVM, IDS can detect abnormal traffic more accurate. There is an innovation in fitness function which is formed from TPR, FPR and the number of selected features. The new fitness function reduced the dimension of the data, increased true positive detection and simultaneously decreased false positive detection. In addition, the computation time for training will also have a remarkable reduction. This study proposes a method which can achieve more stable features in comparison with other techniques. The proposed model has been evaluated test with KDD CUP 99 and UNSW-NB15 datasets. Numeric Results and comparison to other models have been presented.

Handwritten digit recognition is one of the classical issues in the field of image grouping, a subfield of computer vision. The event of the handwritten digit is generous. With a wide opportunity, the issue of handwritten digit recognition by using computer vision and machine learning techniques has been a well-considered upon field. The field has gone through an exceptional turn of events, since the development of machine learning techniques. Utilizing the strategy for Support Vector Machine (SVM) and Principal Component Analysis (PCA), a robust and swift method to solve the problem of handwritten digit recognition, for the Kannada language is introduced. In this work, the Kannada-MNIST dataset is used for digit recognition to evaluate the performance of SVM and PCA. Efforts were made previously to recognize handwritten digits of different languages with this approach. However, due to the lack of a standard MNIST dataset for Kannada numerals, Kannada Handwritten digit recognition was left behind. With the introduction of the MNIST dataset for Kannada digits, we budge towards solving the problem statement and show how applying PCA for dimensionality reduction before using the SVM classifier increases the accuracy on the RBF kernel. 60, 000 images are used for training and 10, 000 images for testing the model and an accuracy of 99. 02% on validation data and 95. 44% on test data is achieved. Performance measures like Precision, Recall, and F1-score have been evaluated on the method used.

This paper is concerned with the development of a novel classifier for automatic mass detection of mammograms, based on contourlet feature extraction in conjunction with statistical and fuzzy classifiers. In this method, mammograms are segmented into regions of interest (ROI) in order to extract features including geometrical and contourlet coefficients. The extracted features benefit from the superiority of the contourlet method to the state of the art multi-scale techniques. A genetic algorithm is applied for feature weighting with the objective of increasing classification accuracy. Although fuzzy classifiers are interpretable, the majority is order sensitive and suffers from the lack of generalization. In this study, a kernel SVM is integrated with a nerofuzzy rule-based classifier to form a support vector based fuzzy neural network (SVFNN). This classifier benefits from the superior classification power of SVM in high dimensional data spaces and also from the efficient human-like reasoning of fuzzy and neural networks in handling uncertainty information. We use the Mammographic Image Analysis Society (MIAS) standard data set and the features extracted of the digital mammograms are applied to the fuzzy-SVM classifiers to assess the performance. Our experiments resulted in 95.6%,91.52%,89.02%, 85.31% classification accuracy for the subclass FSVM, SVFNN, fuzzy rule based and kernel SVM classifiers respectively and we conclude that the subclass fuzzy-SVM is superior to the other classifiers.

Identifying the source camera of an image is one of the most important issues of digital court and is useful in many applications, such as images that are presented in court as evidence. In many methods, the image noise characteristics, extraction of Sensor Pattern Noise and its correlation with non-uniformity of the light response (PNU) are used. In this paper we have presented a method based on photo response non uniformity (PRNU) that provides some features for classification by support vector machine (SVM). Because the noise model is affected by the complexity of the image, we used the wavelet transform to de-noise and reduce edge effects in PRNU noise pattern and also raise the detection accuracy. We also used the Precision processing theory to reduce the image size, then we simplified and summarized the data using the Single Value Decomposition (SVD) Or principal component analysis (PCA). The results show that using two-level wavelet transform and summarized data is more suitable using PCA.

The problem of automatic signature recognition has received little attention in comparison with the problem of signature verification, despite its potential applications for many business processes and can be used effectively in paperless office projects. This paper presents model-based off-line signature recognition with rotation invariant features. Non-linear rotation of signature patterns is one of the major difficulties to be solved in this problem. The proposed system is designed based on support vector machines (SVM) classifier technique and rotation invariant structure feature to tackle the problem. Our designed system consists of three stages: the first is preprocessing stage, the second is feature extraction stage and the last is SVM classifier stage. Experimental results demonstrated that the proposed methods were effective to improve recognition accuracy.