Matches in UGent Biblio for { ?s ?p Texture plays an important role in image processing analysis. It can be defined as the pattern describing a surface due to variations of data at scales smaller than the scale of interest. The study of texture permits to characterize and discriminate regions in images. In some cases objects can be masked by texture and removing the texture facilitates the detection of the object. One of the most important topics of image analysis based on texture is to automate the visual inspection of surface in materials. This type of analysis is commonly performed by human experts offering subjective results and being unhealthy for the inspectors. Therefore an automated inspection system converts traditional subjective inspections into objective ones. Several techniques for two-dimensional texture analysis have been developed, increasing the role of texture analysis in practical industrial applications. Specifically Local Binary Pattern (LBP) techniques have proved to be successful for texture classification and texture characterization in related domains.we extend the LBP technique by describing structures around the given pixel. For this, we propose to evaluate the relationship between the points on the circles with different radii around a central pixel. The connections are binarized and placed into a matrix, which can be evaluated using symmetry to group similar versions of patterns.. }
Showing items 1 to 1 of
1
with 100 items per page.
- aggregation abstract "Texture plays an important role in image processing analysis. It can be defined as the pattern describing a surface due to variations of data at scales smaller than the scale of interest. The study of texture permits to characterize and discriminate regions in images. In some cases objects can be masked by texture and removing the texture facilitates the detection of the object. One of the most important topics of image analysis based on texture is to automate the visual inspection of surface in materials. This type of analysis is commonly performed by human experts offering subjective results and being unhealthy for the inspectors. Therefore an automated inspection system converts traditional subjective inspections into objective ones. Several techniques for two-dimensional texture analysis have been developed, increasing the role of texture analysis in practical industrial applications. Specifically Local Binary Pattern (LBP) techniques have proved to be successful for texture classification and texture characterization in related domains.we extend the LBP technique by describing structures around the given pixel. For this, we propose to evaluate the relationship between the points on the circles with different radii around a central pixel. The connections are binarized and placed into a matrix, which can be evaluated using symmetry to group similar versions of patterns.".