Wyniki wyszukiwania - BazTech

1

Novel visual object descriptor using surf and clustering algorithms

Grycuk R.

Journal of Applied Mathematics and Computational Mechanics

|

2016

|

Vol. 15, nr 3

37--46

EN

In this paper we propose a method for object description based on two wellknown clustering algorithms (k-means and mean shift) and the SURF method for keypoints detection. We also perform a comparison of these clustering methods in object description area. Both of these algorithms require one input parameter; k-means (k, number of objects) and mean shift (h, window). Our approach is suitable for images with a non-homogeneous background thus, the algorithm can be used not only on trivial images. In the future we will try to remove non-important keypoints detected by the SURF algorithm. Our method is a part of a larger CBIR system and it is used as a preprocessing stage.

2

Efektywność wykrywania punktów charakterystycznych w obrazach stereoskopowych za pomocą wartości własnych hesjanu

Korohoda P.

Automatyka / Akademia Górniczo-Hutnicza im. Stanisława Staszica w Krakowie

|

2009

|

T. 13, z. 3/2

1233-1243

PL

W artykule opisano eksperyment obliczeniowy, przeprowadzony dla stereoskopowej pary obrazów, mający na celu zbadanie różnic i podobieństw w efektywności wykrywania punktów charakterystycznych z wykorzystaniem trzech metod bazujących na wartościach własnych hesjanu. Stosując gaussowską czteropoziomową piramidę, wyznaczono punkty charakterystyczne wspólne dla wszystkich poziomów, porównano wizualnie i statystycznie ich lokalizację, a następnie wyznaczono mapy dysparycji. W badaniach wykorzystano korelacyjną technikę wyznaczania dysparycji. Wykazano różnice między teoretycznie podobnymi metodami wykorzystującymi iloraz wartości własnych oraz potwierdzono korzystne cechy metody bazującej na różnicy wartości własnych.

EN

In the paper a computational experiment performed for the stereoscopic image pair has been described. The aim of the designed experiment was to investigate the similarities and differences between three methods used to localize the keypoints, basing on the Hessian matrix eigenvalues. The keypoints were obtained for the four-level Gaussian pyramid, and then the ones common for all levels were selected. Afterwards their locations were visually and statistically compared and then the disparity maps were computed. The disparity in that task was calculated with use of the correlation coefficient. The results indicate noticeable differences between theoretically similar techniques utilizing the eigenvalues ratio. The advantegous properties of the third method, based on the difference between eigenvalues, have been confirmed.

3

Corner-based keypoints for scale-invariant detection of partially visible objects

Śluzek A.

Machine Graphics and Vision

|

2006

|

Vol. 15, No. 3/4

639-647

EN

Local features (also known as interest points, keypoints, etc.) are a popular and powerful tool for matching images and detecting partially occluded objects. While the problems of photometric distortions of images and rotational invariance of the features have satisfactory solutions, satisfactorily simple scale-invariant algorithms do not exist yet. Generally, either computationally complex methods of scale-space (multi-scale approach) are used, or the correct scale is estimated using additional mechanisms. The paper proposes a new category of keypoints that can be used to develop a simple scale-invariant method for detecting known objects in analyzed images. Keypoints are defined as locations at which selected moment-based parameters are consistent over a wide range of different-size circular patches around the keypoint. While the database of known objects (i.e. the keypoints and their descriptions) is still built using a multi-scale approach, analyzed images are scanned using only a single-scale window and its sub-window. The paper focuses on the keypoint building and keypoint matching principles. Higher-level issues of hypotheses building and verification (regarding the presence of objects in analyzed images) are only briefly discussed.