Device and method for analysing images by calculating the variance for each pixel of a high-frequency image
View Patent ↗A method and device for analysing video images by obtaining high frequency components. The device includes means for obtaining high-frequency components of a video image and means for obtaining a high-frequency image on the basis of the high-frequency components. Means for calculating the variance for each pixel of the high-frequency image. Means for associating a coefficient with each pixel of the video image as a function of the variance. The coefficient is representative of the visual sensitivity of each pixel for the human eye. The invention relates to a method and a device for analysing video images.
1. Device for analysing video images wherein it comprises
means of obtaining high-frequency components of a video image comprising means of wavelet transformation of said video image,
means of obtaining a high-frequency image on the basis of said high-frequency components comprising means of weighting the wavelet transformation of said video image, assigning a greater weighting to a high-frequency image representative of diagonal contours,
means of calculating a variance for each pixel of said high-frequency image,
means of associating a coefficient with each pixel of the video image as a function of said variance, said coefficient being representative of a visual sensitivity of each pixel for a human eye.
2. Device according to claim 1 wherein the means of calculating the variance of each pixel define for each pixel a neighbourhood centred on said pixel and calculate a mean deviation of the values of each pixel in a neighbourhood.
3. Device according to claim 1 further comprising means of slicing a scale of the variances into equal length spans, wherein the means of associating said coefficient with each pixel allocates a value to said coefficient per span.
4. Device according to claim 3 wherein the means of associating said coefficient modify said coefficient as a function of a distance of a current pixel from the centre of the video image.
5. Device according to claim 3 further comprising means of detecting zones with weak motion in the video image.
6. Device according to claim 5 wherein the means of associating said coefficient for each pixel, decrease said value of said coefficient to a minimum value when said pixel associated with said coefficient belongs to a fixed zone.
7. Device according to claim 1 , for video pre-processing comprising means of entropy reduction on an incoming video signal, means of performing a mixing operation between said incoming video signal and the video signal after entropy reduction, wherein said coefficient performing a weighting between the incoming video signal and the video signal after entropy reduction in the means of mixing.
8. Device according to claim 1 , for video coding, wherein said coefficient making it possible to adjust a compression rate applied by said device for analysing images.
9. Device according to claim 1 , for digital watermarking further comprising a device for analysing video images, wherein said coefficient making it possible to determine a zone of the image for containing a watermarking.
10. Method of analysing video images wherein it comprises
obtaining high-frequency components of an image comprising means of wavelet transformation of said video image,
obtaining a high-frequency image on a basis of said high-frequency components wherein the wavelet transformation of said video image is weighted by assigning a greater weighting to a high-frequency image representative of diagonal contours,
calculating a variance for each pixel of said high-frequency image,
associating a coefficient for each pixel of the image as a function of said variance, said coefficient being representative of a visual sensitivity of each pixel for a human eye.