Super-resolution reconstructing method for enhancing smoothness and sharpness of video image

摘要

A super-resolution reconstructing method includes: providing an original image including multiple original image pixel points arranged in an array; performing edge detecting on the original image by Canny_H method and considering pixel edge information of 4*4 area image range; performing 2*2 times interpolation zooming to edge pixel points and determining a type of the 4*4 area image as pure color entity area or non-pure color entity area according to the pixel edge information; performing a conditional interpolation individually on pure color entity area and non-pure color entity area; and obtaining a pixel 2*2 times interpolation zoomed super-resolution image. In pure color entity area, interpolation is performed according to edge information of 4*4 area image range to maintain border sharpness of pure color entity area and reduce jaggies. In non-pure color entity area, BiCubic_H interpolation is performed to enhance smoothness and sharpness of image.

主权项

1. A super-resolution reconstructing method for enhancing smoothness and sharpness of video image, comprising:
step 1: providing an original image, wherein the original image comprises a plurality of original image pixel points (P) arranged in an array; step 2: performing an edge detecting to the original image by use of Canny_H method and considering pixel edge information of a 4*4 area image; step 3: performing a 2*2 times interpolation zooming on edge pixel points and determining a type of the area image as pure color entity area or non-pure color entity area according to the pixel edge information; step 4: performing a conditional interpolation individually on pure color entity area and non-pure color entity area; and step 5: obtaining a pixel 2*2 times interpolation zoomed super-resolution image.