The following information was submitted:
Transactions: WSEAS TRANSACTIONS ON INFORMATION SCIENCE AND APPLICATIONS
Transactions ID Number: 27-608
Full Name: Ioan Lita
Position: Professor
Age: ON
Sex: Male
Address: str. Targu din Vale, Nr. 1 Pitesti, Arges, Cod: 110040
Country: ROMANIA
Tel:
Tel prefix:
Fax:
E-mail address: lita@upit.ro
Other E-mails: visan@upit.ro,m_jurian@upit.ro
Title of the Paper: A study for comparative evaluation of the methods for image processing using texture characteristics
Authors as they appear in the Paper: Mariana Jurian,Ioan Lita,Florentina Enescu,Daniel Alexandru Visan
Email addresses of all the authors: m_jurian@yahoo.com,lita@upit.ro,visan@upit.ro
Number of paper pages: 10
Abstract: In this paper a comparative study has been made regarding certain methods of image processing by the texture characteristic, to find the optimum method for detecting color texture. The study is based on the experiments. There are two main points: the quality of the detection and the response time. For the experiments have been used the co-occurrence matrix and the iso-segments matrix of the gray or color levels. These methods are based on processing the image at a pixel level and the making of matrix that contain certain spatial positions of pixels. The resulting matrix was analyzed and, based on the existing information, the characteristics vectors associated to the matrix was determined. Each method was studied by analyzing the pixels of the image files in certain situations and directions and at certain distances. The co-occurrence methods had better results for the queries based on water, wood and grass textures and on the ones based on sand, ruble and clouds, !
better results were obtained using the iso-segment matrix.
Keywords: Texture,Co-occurrence,Iso-segments,Pixel,Grain,Contrast,Directionality,Regularity,Roughness,Line-likeness
EXTENSION of the file: .pdf
Special (Invited) Session: A comparative study of the methods for image processing using texture characteristics
Organizer of the Session: 594-457
How Did you learn about congress:
IP ADDRESS: 92.83.112.213
