A preprocessing approach to ROI video coding using variable Gaussian filters and variance in intensity

In applications involving video over mobile phones or Internet, the limited quality depending on the transmission rate can be further improved by region-of-interest (ROI) coding. In this paper we present a preprocessing method using variable Gaussian filters controlled by a quality map indicating the distance to the ROI border that seeks to smooth the border effects between ROI and non-ROI According to subjective tests the reduction of border effects increases the perceived quality, compared to using only one low pass filter. It also introduces a small improvement of the PSNR of the intensity component within the ROI after compression. With the compressed original sequence as a reference, the average PSNR was increased by 1.25 dB and 2.3 dB for 100 kbit/s and 150 kbit/s, respectively. Furthermore, in order to reduce computational complexity, a modified quality map is introduced using variance in intensity to exclude pixels, which are not visibly affected by the Gaussian filters. No change in quality is noticed when using less than 76% of the pixels