Preliminary results on sky segmentation
- Resource Type
- Conference
- Authors
- Nafornita, Corina; David, Ciprian; Isar, Alexandru
- Source
- 2015 International Symposium on Signals, Circuits and Systems (ISSCS) Signals, Circuits and Systems (ISSCS), 2015 International Symposium on. :1-4 Jul, 2015
- Subject
- Components, Circuits, Devices and Systems
Computing and Processing
Signal Processing and Analysis
Image reconstruction
Noise reduction
Image segmentation
Estimation
Wavelet transforms
Buildings
Cameras
- Language
The goal of this paper is to classify an image into two regions: sky and non-sky. We discuss here the possibility to use a non-linear simplifying filter that operates on high frequencies of the color image, more specifically on the Blue channel, and next to use a k-means classifier with two clusters to perform the sky segmentation.