Research on Image Complexity Description Method Based on Approximate Entropy

Hong Zhang,Jiongwei Dong,Shasha He,Shengze Lv
DOI: https://doi.org/10.1109/icitbs49701.2020.00203
2020-01-01
Abstract:In order to better describe the complexity of images, this paper proposes a method to describe the complexity of images, namely to measure the complexity of images by the approximate entropy value of time series. In order to verify the effectiveness of the method, first, design the image complexity of subjective evaluation experiment, through the subjective ratings experiment get different complexity of images, and then on the basis of general experimental images into one dimensional time series, the complexity of calculating the approximate entropy measures image, and the existing test method is used to describe the complexity of the image. Experimental results show that this method can reflect the overall complexity of images, and can be used for qualitative and quantitative comparison of different image complexity.
What problem does this paper attempt to address?