Showing papers in "Pattern Recognition Letters in 2011"

TL;DR: VSUMM is presented, a methodology for the production of static video summaries that is based on color feature extraction from video frames and k-means clustering algorithm and develops a novel approach for the evaluation of video static summaries.

TL;DR: This paper investigates a simple but powerful approach to make robust use of HOG features for face recognition by proposing to extract HOG descriptors from a regular grid and identifying the necessity of performing dimensionality reduction to remove noise and make the classification process less prone to overfitting.

TL;DR: A linear time line segment detector that gives accurate results, requires no parameter tuning, and runs up to 11 times faster than the fastest known line segment detectors in the literature; hence the name EDLines.

TL;DR: This paper proves that Otsu threshold is equal to the average of the mean levels of two classes partitioned by this threshold, and proposes an improved Otsi algorithm that constrains the search range of gray levels.

TL;DR: The mean curvature method, which views the vein image as a geometric shape and finds the valley-like structures with negative mean curvatures, achieves 0.25% equal error rate, which is significantly lower than what existing methods can achieve.

TL;DR: A novel combination of vision based features in order to enhance the recognition of underlying signs and kurtosis position and principal component analysis, PCA are presented.

TL;DR: An algorithm to compute initial cluster centers for k-means algorithm is proposed and is applied to several different datasets in different dimension for illustrative purposes and it is observed that the newly proposed algorithm has good performance.

TL;DR: Comparisons with the segmentation results of a gradient vector flow deformable (GVF) model and a region based active contour model (ACM) are performed, which indicate that the proposed method produces more accurate nuclei boundaries that are closer to the ground truth.

TL;DR: A new clustering method, called DE-KM, which combines differential evolution algorithm (DE) with the well known K-means procedure is described, which shows that if the number of clusters K is sufficiently large, DE-kM obtains solutions with lower SSE values than the other five algorithms.

TL;DR: In this paper, a word-based off-line recognition system is proposed, using Hidden Markov Models (HMMs), which yields superior results of improved accuracy in comparison with several typical methods.

TL;DR: A homomorphic filtering-based illumination normalization method that is simple and computationally fast because there are mature and fast algorithms for the Fourier transform used in homomorphic filter and the Eigenfaces method is chosen to recognize the normalized face images.

TL;DR: This work proposes a stochastic algorithm based on the GRASP meta-heuristic, with the main goal of speeding up the feature subset selection process, basically by reducing the number of wrapper evaluations to carry out.

TL;DR: Experimental evidence is provided that CC-RANSAC may recover the planar patches composing a typical step or ramp with substantially higher accuracy than the traditional RANSAC algorithm.

TL;DR: This paper proposes an improvement of OS-ELM based on the bi-objective optimization approach, which tries to minimize the empirical error and obtain small norm of network weight vector on benchmark datasets.

TL;DR: The results of the experiments show that the proposed method adapts well to all types of binarization challenges, can deal with higher numbers ofbinarization problems and boosts the overall performance of the binarizations.

TL;DR: Two cluster validity indices are proposed for efficient validation of partitions containing clusters that widely differ in sizes and densities, each based on a compactness measure and a separation measure.

TL;DR: A novel idea based on computer vision is presented and Genetic Algorithm has been used to select features to get the best information for diagnosing the disease.

TL;DR: A local density adaptive similarity measure is proposed, which uses the local density between two data points to scale the Gaussian kernel function and satisfies the clustering assumption and has an effect of amplifying intra-cluster similarity, thus making the affinity matrix clearly block diagonal.

TL;DR: The lesion features used in the classification framework are inspired on border, texture, color and structures used in popular dermoscopy algorithms performed by clinicians by visual inspection, which allows for the detection of particular dermoscopic patterns associated with melanoma.

TL;DR: This paper gives a comprehensive survey and categorisation of computer vision and pattern recognition techniques proposed so far against image spam, and makes an experimental analysis and comparison of some of them on real, publicly available data sets.

TL;DR: A novel clustering ensemble method, SELective Spectral Clustering Ensemble (SELSCE), is proposed and the experimental results demonstrate that the proposed algorithm can achieve a better result than the traditional clusteringsemble methods.

TL;DR: This study aims at tapping the potential of t-norms for multimodal biometrics as it is quite computationally fast and outperforms the score level fusion using the combination approach (min, mean, and sum) and classification approaches like SVM, logistic linear regression, MLP, etc.

TL;DR: An approximate run length based scheme is proposed to detect image splicing and demonstrates that the proposed approach can achieve a relatively high accuracy with less computational cost and fewer features when compared with other methods.

TL;DR: This paper presents two content-based image retrieval frameworks with relevance feedback based on genetic programming that outperformed six other relevance feedback methods regarding their effectiveness and efficiency in image retrieval tasks.

TL;DR: A novel method for accurate and fast computation of orthogonal Gegenbauer moments is proposed and the efficiency and the superiority of the proposed method are explained.

TL;DR: A new approach for edge detection using a combination of bacterial foraging algorithm (BFA) and probabilistic derivative technique derived from Ant Colony Systems, is presented in this paper.

TL;DR: A new convolutional neural network architecture that includes the fast Fourier transform between two hidden layers to switch the signal analysis from the time domain to the frequency domain inside the network is presented.

TL;DR: This paper proposes a new semi-supervised constraint score that uses both pairwise constraints and local properties of the unlabeled data and shows that this new score is less sensitive to the given constraints than the previous scores while providing similar performances.

TL;DR: A gradient similarity-matching measure was implemented in a state-of-the-art local stereo- matching method (an adaptive support-weight algorithm) and it was the best local area-based method compared to the permanent Middlebury table entries.

TL;DR: Experimental results show that Gaussian noise added to low-quality fingerprint images enables the extraction of useful features for biometric identification by adding noise to the original signal.