Search

This paper presents a methodology for extracting road seeds (segments) from digital aerial images, these images often have variations, which involves complex analysis and characterizes a challenging problem in this area. This work consists in implementing techniques for detecting possible road seed. This paper presents a methodology in eliminating invalid points of the roads using the application of Modified Canny Algorithm. In the segmentation of valid points for seed construction was applied the Douglas-Peucker algorithm. In the road extraction, problems were found in relation to valid and invalid seed and in images that have rural roads. Experiments were performed with aerial images of roads to prove the proposed methodology. [ABSTRACT FROM AUTHOR]

This paper presents a method for building a three dimensional model of the lower leg bone (tibia) from a human, from X-ray images. The methodology adopted to obtain the 3D object is deforming the model until it corresponds to X-ray images (front and left and right sides) of the tibia. We used the following algorithms for X-ray images segmentation and three-dimensional modeling: Canny, Border Following, Free Form Deformation (FFD) and deformation algorithm developed in this work. The viewing application defined and implemented shows the three-dimensional image of the bone of the tibia resulting from deformation of the 3D object. The experimental results demonstrate the possibility of representing the real human bone obtained from X-ray images into a 3D model, however, a great difficulty for this activity was the number of images available for the project. [ABSTRACT FROM AUTHOR]