Silhouette recognition using high-resolution pursuit

Jaggi, Sidharth; Karl, W.C.; Mallat, Stéphane; Willsky, Alan S.

doi:10.1016/s0031-3203(98)00112-5

Cited by 15 publications

(11 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Both local and global methods have been explored for automatic visual object recognition (to be distinguished from object detection, which suggests different methods). Local methods use features such as critical points (Freeman, 1978) or high-resolution pursuit (Jaggi et al, 1999). Systems using local features can perform well in the presence of noise, distortion, or partial occlusion because only one distinctive part need be recognized.…”

Section: Introductionmentioning

confidence: 99%

Robust Recognition of Ship Types from an Infrared Silhouette

Alves¹,

Herman²,

Rowe

2004

View full text Add to dashboard Cite

Accurate identification of unknown contacts crucial in military intelligence. Automated systems that quickly and accurately determine the identity of a contact could be a benefit in backing up electronic-signals identification methods. This work reports two experimental systems for ship classification from infrared FLIR images. In an edge-histogram approach, we used the histogram of the binned distribution of observed straight edge segments of the ship image. Some simple tests had a classification success rate of 80% on silhouettes. In a more comprehensive neuralnetwork approach, we calculated scale-invariant moments of a silhouette and used them as input to a neural network. We trained the network on several thousand perspectives of a wire-frame model of the outline of each of five ship classes. We obtained 70% accuracy with detailed tested on real infrared images but performance was more robust than with the edge-histogram approach.

show abstract

Section: Introductionmentioning

confidence: 99%

Robust Recognition of Ship Types from an Infrared Silhouette

Alves¹,

Herman²,

Rowe

2004

View full text Add to dashboard Cite

show abstract

“…Step1: Obtain a silhouette image and extract 1D representation of the silhouette image using centroidal distance profile (CDP) [35], [36]. 2) Step2: Divide the CDP into overlapping intervals that are unwrapped segments of the boundary curve.…”

Section: )mentioning

confidence: 99%

The Relevance Sample-Feature Machine: A Sparse Bayesian Learning Approach to Joint Feature-Sample Selection

Mohsenzadeh

Sheikhzadeh

Azimi

et al. 2013

IEEE Trans. Cybern.

View full text Add to dashboard Cite

Abstract-This paper introduces a novel sparse Bayesian machine-learning algorithm for embedded feature selection in classification tasks. Our proposed algorithm, called the relevance sample feature machine (RSFM), is able to simultaneously choose the relevance samples and also the relevance features for regression or classification problems. We propose a separable model in feature and sample domains. Adopting a Bayesian approach and using Gaussian priors, the learned model by RSFM is sparse in both sample and feature domains. The proposed algorithm is an extension of the standard RVM algorithm, which only opts for sparsity in the sample domain. Experimental comparisons on synthetic as well as benchmark data sets show that RSFM is successful in both feature selection (eliminating the irrelevant features) and accurate classification. The main advantages of our proposed algorithm are: less system complexity, better generalization and avoiding overfitting, and less computational cost during the testing stage.Index Terms-Joint feature selection and classifier design, relevance features, relevance sample feature machine (RSFM), relevance samples, sparse Bayesian learning.

show abstract

“…Mu ltiple techniqu es are u sed in sketch recognition to d etect or classify regu lar geom etric shapes [6][7][8], hand w riting characters [9,10], fingerprints [11], electric circu its [12,13], d iagram s [14,15], and other u ser com m and gestu res. For instance, w ith a classic linear d iscrim inator, Rubine [16] calcu lated featu res in ord er to classify single-stroke sketches as d igits, letters and basic com m and s introd u ced in a specific w ay.…”

Section: Related Workmentioning

confidence: 99%

New method to find corner and tangent vertices in sketches using parametric cubic curves approximation

Albert

Fernández-Pacheco

2013

Pattern Recognition

View full text Add to dashboard Cite

Aleixos Borrás, MN. (2013). New method to find corner and tangent vertices in sketches using parametric cubic curves approximation. Pattern Recognition. 46 (5) Abstract-Some recent approaches have been presented as simple and highly accurate corner finders in the sketches including curves, which is useful to support natural human-computer interaction, but these in most cases do not consider tangent vertices (smooth points between two geometric entities, present in engineering models), what implies an important drawback in the field of design. In this article we present a robust approach based on the approximation to parametric cubic curves of the stroke for further radius function calculation in order to detect corner and tangent vertices. We have called our approach Tangent and Corner Vertices Detection (TCVD), and it works in the following way.First, corner vertices are obtained as minimum radius peaks in the discrete radius function, where radius is obtained from differences. Second, approximated piecewise parametric curves on the stroke are obtained and the analytic radius function is calculated. Then, curves are obtained from stretches of the stroke that have a small radius.Finally, the tangent vertices are found between straight lines and curves or between curves, where no corner vertices are previously located. The radius function to obtain curves is calculated from approximated piecewise curves, which is much more noise free than discrete radius calculation. Several tests have been carried out to compare our approach to that of the current best benchmarked, and the obtained results show that our approach achieves a significant accuracy even better finding corner vertices, and moreover, tangent vertices are detected with an Accuracy near to 92% and a False Positive Rate near to 2%.

show abstract

Silhouette recognition using high-resolution pursuit

Cited by 15 publications

References 24 publications

Robust Recognition of Ship Types from an Infrared Silhouette

Robust Recognition of Ship Types from an Infrared Silhouette

The Relevance Sample-Feature Machine: A Sparse Bayesian Learning Approach to Joint Feature-Sample Selection

New method to find corner and tangent vertices in sketches using parametric cubic curves approximation

Contact Info

Product

Resources

About