Method of image super-resolution reconstruction

Abstract

The invention discloses a method image super-resolution reconstruction. The method comprises steps of processing a high-resolution image to obtain an interpolation image of a low-resolution image; training to obtain a high and low resolution dictionary pair; inputting the low-resolution image and carrying out interpolation processing to obtain an interpolation image of the low-resolution image; decomposing the interpolation image of the low-resolution image into a low-resolution structure part and a low-resolution texture part, and abandoning the low-resolution texture part; extracting features of the low-resolution interpolation image to obtain low-resolution image features; according to the high and low resolution dictionary pair, carrying out sparse reconstruction on the low-resolution image features to obtain a high-resolution image texture part; and combining the high-resolution image texture part and the low-resolution structure part to obtain a reconstructed high-resolution image. According to the invention, corresponding samples are specifically classified and trained, and the corresponding dictionary pair is classified, trained and then used for following super-resolution reconstruction, so resolution of a reconstructed image can be precisely improved.

Description

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine

Login to View More

AI Technical Summary

Benefits of technology

This technology allows us to combine two different types of data - one very detailed but another less precise than what we are looking for or have previously done by analyzing both kinds of data separately. By doing this, our algorithm improves accuracy when combining these two sources of data together without affecting their quality.

Problems solved by technology

This patented technical problem addressed in this patents relates to improving image quality when analyzing large amounts of pixel values without losing any detail or noise caused by factors like sensors size limitations. Existing methods involve performing interpolation between different dictionaries with varying levels of accuracy but they only work well if all previous ones were accurate enough.

Images