Institute of Information Theory and Automation

You are here

Effective image restoration algorithms for mobile devices

Jan Kamenický
Defense type: 
Date of Event: 
FJFI ČVUT, Trojanova 13, Praha 2, místnost 212
Current mobile devices are usually equipped with low-budget cameras with low quality optics and sensors. Therefore, digital images acquired by these cameras are generally deteriorated by noise and blur and have effective resolution lower than the number of pixels. The task of image restoration is the recovery of a noise-free and sharp image with possibly higher spatial resolution from one or several such input images. For achieving all these tasks together, an image restoration framework is presented which performs registration, denoising, deblurring and resolution enhancement. The main task is however the effective and fast implementation of the algorithms in special scenarios. The first is the image quality improvement of mobile phone cameras by means of multichannel blind deconvolution and another one is the resolution enhancement of infra-red video using super-resolution algorithms. Real working applications for these scenarios used in practise are the main contribution of this work. Implementation details are discussed together with results of thorough testing on real data.
2018-05-03 08:01