Narrow Results

Learning vector quantization (LVQ) network and back-propagation (BP) network are constructed easily making use of MATLAB toolbox on the basis of maintaining the recognition rate. Face images are randomly selected from images set as training data of LVQ network and BP network. LVQ algorithm and BP algorithm are used to train network. The automatic recognition of face orientation is realized when the system obtains convergence network. First, all images are processed by edge detection. Then feature vectors representing position of the eye were extracted from edge detected images. Feature vectors of training set are sent to network to adjust the parameters which ensures the convergence speed and performance of the network. Experimental results show that the constructed LVQ network and BP network can judge face orientation according to feature vectors of input images. Generally, the recognition rate of LVQ network is higher than that of BP network. The LVQ network and BP network are both feasible and effective for face orientation recognition to some extent. The advantage of this work is that the recognition system is efficient and easy to promote. This paper focuses on how to use MATLAB easily to design identification network rather than the complexity of identification system. The future research will focus on the stability and robustness of recognition network.

The fractal geometries are applied extensively in many applications like pattern recognition, texture analysis and segmentation. The application of fractal geometry requires estimation of the fractal features. The fractal dimension and fractal length are found effective to analyze and measure image features, such as texture, resolution, etc. This paper proposes a new wavelet–fractal technique for image resolution enhancement. The resolution of the wavelet sub-bands are improved using scaling operator and then it is transformed into texture vector. The proposed method then computes fractal dimension and fractal length in gradient domain which is used for resolution enhancement. It is observed that by using scaling operator in the gradient domain, the fractal dimension and fractal length becomes scale invariant. The major advantage of the proposed wavelet–fractal technique is that the feature vector retains fractal dimension and fractal length both. Thus, the resolution enhanced image restores the texture information well. The texture information has also been observed in terms of fractal dimension with varied sample size. We present qualitative and quantitative analysis of the proposed method with existing state of art methods.

In response to the fast propagation of videos on the Internet, Content-Based Video Retrieval (CBVR) was introduced to help users find their desired items. Since most videos concern humans, human action retrieval was introduced as a new topic in CBVR. Most human action retrieval methods represent an action by extracting and describing its local features as more reliable than global ones; however, these methods are complex and not very accurate. In this paper, a low-complexity representation method that more accurately describes extracted local features is proposed. In this method, each video is represented independently from other videos. To this end, the motion information of each extracted feature is described by the directions and sizes of its movements. In this system, the correspondence between the directions and sizes of the movements is used to compare videos. Finally, videos that correspond best with the query video are delivered to the user. Experimental results illustrate that this method can outperform state-of-the-art methods.

In this paper, an image clustering method that is essential for content-based image retrieval in large image databases efficiently is proposed by color, texture, and shape contents. The dominant triple HSV (Hue, Saturation, and Value), which are extracted from quantized HSV joint histogram in the image region, are used for representing color information in the image. Entropy and maximum entry from co-occurrence matrices are used for texture information and edge angle histogram is used for representing shape information. Due to its algorithmic simplicity and the several merits that facilitate the implementation of the neural network, Fuzzy ART has been exploited for image clustering. Original Fuzzy ART suffers unnecessary increase of the number of output neurons when the noise input is presented. Therefore, the improved Fuzzy ART algorithm is proposed to resolve the problem by differently updating the committed node and uncommitted node, and checking the vigilance test again. To show the validity of the proposed algorithm, experimental results on image clustering performance and comparison with original Fuzzy ART are presented in terms of recall rates.

A new feature descriptor called local bit plane-based dissimilarities and adder pattern (LBPDAP) is proposed in this paper for content-based computed tomography (CT) image retrieval. To compute the descriptor, the bit planes of the input image are first extracted. For each pixel of an image, these bit planes are then locally encoded using an adder which combines the center-neighbor dissimilarity information and the neighbor–neighbor mutual dissimilarity information in each bit plane. The encoded bit plane values corresponding to each center pixel are finally compared with the intensity of the center pixel to compute the proposed LBPDAP. In order to limit the feature dimensions, we have considered only four most significant bit planes for LBPDAP computations as the higher bit planes contain more significant visual texture information. The proposed descriptor is low dimensional and experimental results on widely accepted NEMA and TCIA-CT image databases demonstrate better retrieval efficiency of LBPDAP over many recent local pattern-based approaches.

These days identification of a person is an integral part of many computer-based solutions. It is a key characteristic for access control, customized services, and a proof of identity. Over the last couple of decades, many new techniques were introduced for how to identify human faces. This approach investigates the human face identification based on frontal images by producing ratios from distances between the different features and their locations. Moreover, this extended version includes an investigation of identification based on side profile by extracting and diagnosing the feature sets with geometric ratio expressions which are calculated into feature vectors. The last stage involves using weighted means to calculate the resemblance. The approach considers an explainable Artificial Intelligence (XAI) approach. Findings, based on a small dataset, achieve that the used approach offers promising results. Further research could have a great influence on how faces and face-profiles can be identified. Performance of the proposed system is validated using metrics such as Precision, False Acceptance Rate, False Rejection Rate, and True Positive Rate. Multiple simulations indicate an Equal Error Rate of 0.89. This work is an extended version of the paper submitted in ACIIDS 2020.

This paper considers the problem of enumerating all non-isomorphic tree-like chemical graphs with given path frequency, where "tree-like" means that the graph can be viewed as a tree if multiple edges (i.e., edges with the same end points) and a benzene ring are treated as one edge and one vertex, respectively, and "path frequency" is a vector of the numbers of specified vertex-labeled paths that must be realized in every output. This and related problems have several potential applications such as classification of chemical compounds, structure determination using mass-spectrum and/or NMR and design of novel chemical compounds.

For this problem, several studies have been done. Recently, Fujiwara et al. (2008) showed two formulations and for each of them, they gave a branch-and-bound algorithm, which combined efficient enumeration of non-isomorphic trees with bounding operations based on the path frequency and the atom-atom bonds to avoid the generation of invalid trees. In this paper, based on their work and a result of Nagamochi (2006), we introduce two new bounding operations, the detachment-cut and the H-cut, to further reduce the size of the search space. We performed computational experiments to compare our proposed algorithms with those of Fujiwara et al. (2008) using some chemical compound data obtained from the KEGG LIGAND database (http://www.genome.jp/kegg/ligand.html). The results show that our proposed algorithms are much faster than their algorithms.

Fingerprint matching is a key technique of automatic fingerprint recognition system. The result of point pattern matching is accurate while its repeated search for maximum likelihood pairs are very time-consuming. In this paper, we proposed a fast complex fingerprint matching method to solve this problem. The innovative method constructs rotation-and-translation-invariant feature vectors to describe local information of minutiae sets and shifts the matching style from point pattern to feature vector. If the decision is uncertain by this case, then it estimates rotation and translation parameters from previous matched pairs and returns to point matching. Since this method takes the similarity and consistency of local deformation into account, it is not only much faster, but also lower false accepted than traditional approaches. The test on FVC2004 data sets shows that the equal error rate (EER) is 2.97%, and the mean matching time is 10.3 milliseconds, variance is 2.03, superior to most traditional methods. When minutiae number is among 15-40, the EER reaches to 2.14%.

Naïve Bayes is one of simple and practical methods to process classification. But there are two problems. One is that Naïve Bayes can not process incremental classification, and the other is that the time complexity of mass data algorithms is very high. We proposed one method called Incremental Bayes Classification Based on Feature Vector to solve these two problems. And the way to gain feature vector was given through combining vector principle with Euclid Distance. The procedure of incremental classification was described elaborately. At last, the ability of incremental classification of our algorithm was proved better, and the time complexity was lower through doing experiments. At the same time, we analyzed efficiency of our algorithm.

We present a soft computing approach to character recognition from printed documents. For feature extraction, we define a number of fuzzy sets on the Hough transform of character pattern pixels and synthesize additional fuzzy sets by t-norms. The height of each t-norm constitutes a feature element and a set of 'n' such feature elements form an n-dimensional feature vector for the character. A 3n-dimensional vector is then generated from the n-dimensional feature vector by defining three linguistic fuzzy sets, namely, weak, moderate and strong. These 3n-dimensional vectors form a Multilayer Perceptron (MLP) input for training by back propagation. The MLP outputs represent fuzzy sets denoting the belongingness of an input pattern to a number of fuzzy pattern classes. The feature set is chosen by optimizing a Feature Quality Index (FQI) using genetic algorithm. A two-state Markov chain is used to model degraded document images for simulation tests. The system can recognize characters with an accuracy of 98%.

Power quality disturbance identification and classification are important for power quality management, for which a new method is proposed to deal with power quality disturbance identification and classification based on GA-ELM (Genetic Algorithm-Extreme Learning Machine). Decompose the disturbance signals with wavelet for ten layers, it is proposed that using standard deviation and mean of the energy difference, and the ratio of them as the feature vectors. In addition, the root mean of disturbance signals and normal signals is calculated as a supplement in order to reduce the dimension of the importing vectors. The Input Layer and the Hidden Layer Connection Weights and Hidden Layer Neuron Threshold which are set randomly in ELM can influence classification accuracy. So it is proposed that ELM training error is used as the fitness function of GA to optimize Input Layer and the Hidden Layer Connection Weights and Hidden Layer Neuron Threshold to enhance the accuracy of classification and also maintain high classification speed. Simulation results demonstrate that this method can accurately and effectively identify seven common disturbance types.