BOVW FOR CLASSIFICATION IN GEOMETRICS SHAPES

Journal Title: Applied Computer Science - Year 2018, Vol 14, Issue 4

Abstract

The classification of forms is a process used in various areas, to perform a classification based on the manipulation of shape contours it is necessary to extract certain common characteristics, it is proposed to use the bag of visual words model, this method consists of three phases: detection and extraction of characteristics, representation of the image and finally the classification. In the first phase of detection and extraction the SIFT and SURF methods will be used, later in the second phase a dictionary of words will be created through a process of clustering using K-means, EM, K-means in combination with EM, finally in the Classification will be compared algorithms of SVM, Bayes, KNN, RF, DT, AdaBoost, NN, to determine the performance and accuracy of the proposed method.

Authors and Affiliations

Baldemar ZURITA, Luís LUNA, José HERNÁNDEZ, Federico RAMÍREZ

Keywords

Related Articles

AN EFFECTIVE METAHEURISTIC FOR TOURIST TRIP PLANNING IN PUBLIC TRANSPORT NETWORKS

The Time-Dependent Orienteering Problem with Time Windows (TDOPTW) is a combinatorial optimization problem defined on graphs. Its real life applications are particularly associated with tourist trip planning in trans-por...

NUMERICAL ANALYSIS OF THE DRAG COEFFICIENT OF A MOTORCYCLE HELMET

The paper discusses a numerical investigation, using a CFD tool, ANSYS FLUENT, of drag acting on a motorcycle helmet. The simulations were performed on a model of a helmet downloaded from a free CAD model library. A soli...

VERIFICATION OF ACCURACY AND COST OF USE METHODS OF QUALITYASSESSMENT OF IMPLEMENTATION OFDESIGN PATTERNS

Professional programmers use many additional tools over the Integrated Development Environment during their work. Very often they are looking for new solutions, while expecting that the new tool will provide accurate res...

COMPUTER SUPPORT OF ERGONOMIC ANALYSIS OF WORKING CONDITIONS AT WORKSTATIONS

The article shows issues of using computer-aided technologies for ergono-mic analysis at worksplaces using a dedicated tool in the Delmia environment. The scope of the ergonomic evaluation has been limited to the analysi...

IDENTIFICATION OF A BACKLASH ZONE IN AN ELECTROMECHANICAL SYSTEM CONTAINING CHANGES OF A MASS INERTIA MOMENT BASED ON A WAVELET–NEURAL METHOD

In this article a new method of identification of a backlash zone width in a structure of an electromechanical system has been presented. The results of many simulations in a tested model of a complex electromechanical s...

Download PDF file
  • EP ID EP558464
  • DOI 10.23743/acs-2018-25
  • Views 101
  • Downloads 0

How To Cite

Baldemar ZURITA, Luís LUNA, José HERNÁNDEZ, Federico RAMÍREZ (2018). BOVW FOR CLASSIFICATION IN GEOMETRICS SHAPES. Applied Computer Science, 14(4), 5-11. https://europub.co.uk/articles/-A-558464