Tone mapping image quality evaluation method based on gradient magnitude similarity

Abstract

The invention provides a tone mapping image quality evaluation method based on gradient magnitude similarity, and belongs to the field of image processing. The method provided by the invention comprises the following steps: 1) calculating gradient maps of a high dynamic image and a tone mapping image converted by a tone mapping device; 2) carrying out dynamic range adjustment based on a human visual system on the gradient maps corresponding to the high dynamic image and the tone mapping image; 3) calculating the similarity of the gradient maps, and using a tone mapping amplitude as a weight; 4) calculating a natural value of the tone mapping image by means of contrast, brightness and detail description; and 5) merging the similarity value with the natural value. The method provided by the invention is reasonable in design, and can be used for effectively evaluating the quality of the tone mapping image by adopting the gradient magnitude similarity feature capable of effectively capturing image distortion and further referring to the natural features (the image must looknatural) of the image, and thus the evaluation performance is improved.

Description

technical field [0001] The invention belongs to the technical field of image quality evaluation, in particular to a tone mapping image quality evaluation method based on gradient magnitude similarity. Background technique [0002] A high dynamic range image (High Dynamic Range, HDR for short) can provide more dynamic range and image details than an ordinary image, therefore, an HDR image becomes more and more popular. However, common display devices can only display low dynamic range (Low Dynamic Range, LDR for short) images. In order to display the LDR image normally, it needs to be converted into an LDR image. This conversion process is called tone mapping (tonemapping). In recent years, many tone mappers (TMO for short) have emerged. Due to the change of the dynamic range of the image, the evaluation of the converted image quality becomes very important. The evaluation of the quality of tone-mapped images can be used to evaluate the quality of the tone mapper, and can...

AI Technical Summary

Problems solved by technology

Based on the success of feature similarity, the FSITM method based on feature similarity appeared (H.ZiaeiNafchi, A.Shahkolaei, R.FarrahiMoghaddam, and M.Cheriet, "Fsitm: Afeaturesimilarityindexfortone-mappedimages," SignalProcessingLetters, IEEE, vol.22, no .8,pp.1026–1029,2015.), this method uses a locally weighted phase map, but the time complexity of this method is large

[0004] In summary, the existing tone mapping image quality evaluation methods generally have problems such as low evaluation accuracy and complex calculations.

Images