Image enlargement method and image enlargement device

Abstract

The present invention provides an image magnification method and an image magnification device. The image magnification method comprises the following steps: obtaining an original image having a first resolution; performing interpolation magnification on the original image by means of a preset first interpolation algorithm, so as to obtain a first transitional image having a second resolution, the second resolution being greater than the first resolution; performing interpolation magnification on the original image by means of a preset second interpolation algorithm, and smoothing the interpolated and magnified image so as to obtain a second transitional image having the second resolution; performing edge detection on the original image so as to obtain edge information of the original image; building a weight output model, and inputting the edge information of the original image into the weight output model so as to generate a fusion weight of a target image; and fusing the first transitional image and the second transitional image according to the fusion weight and a preset fusion function, so as to obtain the target image having the second resolution. The invention enables smooth transitions at the edges of images, improves image magnification effects, and reduces image magnification costs.

Description

technical field [0001] The invention relates to the field of display technology, in particular to an image magnification method and an image magnification device. Background technique [0002] With the rapid development of computer technology and modern communication technology, people's demand for image information is becoming more and more urgent as human society enters the information age. The current digital image input devices generate corresponding pixel points by sampling small areas on the image, thereby forming a bit-matrix image data, that is, for fixed image input conditions and fixed images, the data that can be obtained The quantity is relatively fixed. [0003] With the development of technology and changes in market demand, consumers have higher and higher requirements for the display quality of display devices. Correspondingly, the resolution of display devices is also getting higher and higher, and the demand for high-resolution video and signals is also in...

AI Technical Summary

Problems solved by technology

[0004] At present, the commonly used image magnification method is generally interpolation magnification. Typical interpolation magnification methods include nearest neighbor interpolation, bilinear interpolation, bicubic interpolation and polynomial interpolation. Among them, the nearest neighbor interpolation algorithm is the simplest, but the nearest neighbor interpolation algorithm is also It is most likely to produce discontinuous pixel values, which will lead to block effects, and then cause image blurring. After zooming in, the image quality effect is generally not ideal.

The bilinear interpolation algorithm is relatively complicated, and the bilinear interpolation algorithm will not cause discontinuous pixel values. The quality of the enlarged image is high, but it will change the edge outline and details of each subject in the image to a certain extent. fuzzy, and the algorithm of bicubic interpolation and polynomial interpolation is more complicated

[0005] Furthermore, in the prior art, the same algorithm is usually used for interpolation and amplification on the flat area and edge area of ​​the image. When a simpler algorithm is selected for operation, there is often obvious jaggedness or distortion in the final image; Although image distortion can be avoided when using a more complex algorithm, the entire calculation process takes a long time and has higher requirements for hardware, which cannot take into account both the effect of image amplification and the cost of amplification.

Images