Showing papers in "Pattern Recognition in 1997"

TL;DR: AUC exhibits a number of desirable properties when compared to overall accuracy: increased sensitivity in Analysis of Variance (ANOVA) tests; a standard error that decreased as both AUC and the number of test samples increased; decision threshold independent; and it is invariant to a priori class probabilities.

TL;DR: These processes and a set of tools to facilitate content-based video retrieval and browsing using the feature data set are presented in detail as functions of an integrated system.

TL;DR: This paper presents a general technique for thresholding of digital images based on Renyi's entropy, which includes two of the previously proposed well known global thresholding methods.

TL;DR: A color segmentation algorithm which combines region growing and region merging processes to generate a non-partitioned segmentation of the image being processed in spatially disconnected but colorimetrically similar regions.

TL;DR: A new clustering algorithm called Competitive Agglomeration (CA), which minimizes an objective function that incorporates the advantages of both hierarchical and partitional clustering, is presented.

TL;DR: A feature-based segmentation approach to the object detection problem is pursued, where the features are computed over multiple spatial orientations and frequencies, which helps in the detection of objects located in complex backgrounds.

TL;DR: This paper describes an approach for integrating a large number of context-dependent features into a semi-automated tool that provides a learning algorithm for selecting and combining groupings of the data, where groupings can be induced by highly specialized features.

TL;DR: A novel data clustering algorithm is described, which is a hybrid approach combining a genetic algorithm with the classical c-means clustering algorithms, and it is shown that substantial improvement of image quality is obtained by using the genetic approach.

TL;DR: Based on 2-D cluster approach, a fast algorithm for point pattern matching is proposed to effectively solve the problems of optimal matches between two point pattern under geometrical transformation and correctly identify the missing or spurious points of patterns.

TL;DR: Different strategies that can be applied in this context to reach a decision (e.g. assignment to a class or rejection), provided the possible consequences of each action can be quantified are examined.

TL;DR: This paper looks at shot detection and characterization using compressed video data directly and proposes a scheme consisting of comparing intensity, row, and column histograms of successive I frames of MPEG video using the chi-square test.

TL;DR: An approach of selecting shape points and outer-layer used for erosion during each iteration of parallel thinning is introduced and the approach produces good skeleton for different types of corners.

TL;DR: The proposed approach utilizes a set of high-frequency channel energies to characterize texture features, followed by a multi-thresholding technique for coarse segmentation, and the number of texture classes is determined by an inter-scale fusion in which the segmentation results at multiple scales are integrated.

TL;DR: The role of signature shape description and shape similarity measure is discussed in the context of signature recognition and verification and the proposed method allows definite training control and at the same time significantly reduces the number of enrollment samples required to achieve a good performance.

TL;DR: Two segmentation algorithms are presented and edge detection and region growing approaches are combined to find large and crisp segments for coarse segmentation towards other applications like object recognition and image understanding.

TL;DR: To assist human analysis of video data, a technique has been developed to perform automatic, content-based video indexing from object motion to analyse the semantic content of the video.

TL;DR: Experimental results indicate that watershed transformation with the algorithms proposed in this paper produces meaningful segmentations, even without a region merging step, which can efficiently improve segmentation accuracy and significantly reduce the computational cost of watershed-based image segmentation methods.

TL;DR: Experimental results show that combination of the classifiers increases reliability of the recognition results and is the unique feature of this work.

TL;DR: The generalized Dunn's index and the Davies-Bouldin index for cluster validation using graph structures, such as GG, RNG and MST are generalized and superiority over some existing cluster validity indices is established.

TL;DR: This paper casts the optimisation process into a Bayesian framework by exploiting the recently reported global consistency measure of Wilson and Hancock as a fitness measure, and demonstrates empirically that the method possesses polynomial convergence time and that the convergence rate is more rapid than simulated annealing.

TL;DR: Several algorithms for preprocessing, feature extraction, pre-classification, and main classification, and modified Bayes classifier and subspace method for the robust main classification are experimentally compared to improve the recognition accuracy of handwritten Japanese character recognition.

TL;DR: It is shown that, whilst both methods are capable of determining cluster validity for data sets in which clusters tend towards a multivariate Gaussian distribution, the parametric method inevitably fails for clusters which have a non-Gaussian structure whilst the scale-space method is more robust.

TL;DR: This work presents two new algorithms for the detection of circles and ellipses which use the FHT algorithm as a basis: Fast Circle Hough Transform (FCHT) and Fast Ellipse Houghtransform (FEHT).

TL;DR: Thresholding, the problem of pixel classification is attempted here using fuzzy clustering algorithms, using segmented regions are fuzzy subsets, with soft partitions characterizing the region boundaries.

TL;DR: This paper addresses automatic interpretation of images of outdoor scenes by using a large database of ground-truth labelled images, a neural network is trained as a pattern classifier, thereby enabling image databases to be queried on scene content.

TL;DR: Empirical investigations show that the proposed MLP-based scheme is superior to the other schemes implemented.

TL;DR: A computationally efficient procedure for skew detection and text line position determination in digitized documents, which is based on the cross-correlation between the pixels of vertical lines in a document, which provides good and accurate results while it requires only a short computational time.

TL;DR: An improved method for eye-feature extraction, descriptions, and tracking using deformable templates using region-based template deformation, which avoids problems such as template shrinking, adjusting the weights of energy terms, failure of orientation adjustment due to some exceptional cases.

TL;DR: The method can estimate qualitatively camera pan, tilt, zoom, roll, and horizontal and vertical tracking and can distinguish pan from horizontal tracking, and tilt from vertical tracking.

TL;DR: From the double-loop type down to the arch type in the order given above, the framework employs both a geometric grouping and a global geometric shape analysis of fingerprint ridges to accomplish the required task.