Title: Layout detection using computer vision

Authors: Nayan N. Shende; S.P. SyedIbrahim

Addresses: Vellore Institute of Technology, Vandalur – Kelambakkam Road, Chennai, Tamil Nadu 600127, India ' Vellore Institute of Technology, Vandalur – Kelambakkam Road, Chennai, Tamil Nadu 600127, India

Abstract: Remote sensing technology has opened the possibility of performing large object detection from satellite imagery. Computer vision which is concentrate on the theory and technology for building AI systems that extract features from images. Layout designing is as interesting yet complicated part in an architectural design. Architectural design is particularly focused about artistic and usability quality of a layout, which is hard to define formally. Layout detection is also a similar concept where such buildings, trees and roads objects are extracted from the satellite images. So as to collect the coordinates from the extracted objects and replicate the same on another parcels. Here we are using two different approaches to extract the objects such as for roads canny edge algorithm and model-based object detection. Similarly for buildings and trees pixel-based object detection and using tensor flow CNN model. By using the entire algorithm, we will be replicating layout in parcels.

Keywords: computer vision; layout detection; Canny edge detection; road detection; building detection; tensor flow CNN model; object detection.

DOI: 10.1504/IJCCIA.2019.103752

International Journal of Computational Complexity and Intelligent Algorithms, 2019 Vol.1 No.2, pp.165 - 177

Received: 20 Mar 2018
Accepted: 23 May 2018
Published online: 26 Nov 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article