Method and device for estimating image blurring degree

Abstract

The invention discloses a method and device for estimating the image blurring degree. The method includes the steps that an image is obtained, then the image is converted, and multi-scale representation of the image is obtained; error vectors between gradient normalization histograms and an original gradient normalization histogram of the image under all scales are calculated, and error vectors under all the scales are weighted and summed with reciprocals of quadratic sums of the error vectors; the estimated value of the blurring degree is obtained based on the error vectors weighted and summed under all the scales. According to the method and device for estimating the image blurring degree in the technical scheme in the embodiment, the good evaluating effect on blurring images with different kinds of content is achieved, and the evaluation result and the human-eye evaluation result have the high correlation.

Description

technical field [0001] The present application relates to image quality evaluation, in particular to a reference-free image blur estimation method and device. Background technique [0002] In various imaging systems, due to factors such as inaccurate focus, relative motion between the system and the object, and detector crosstalk, the image will be blurred, and the image quality will often degrade. Recognition and understanding have a certain impact. The severity of image blur can be quantitatively estimated by a certain evaluation algorithm, which can be used in image quality monitoring and automatic focusing of imaging systems, and can also guide the improvement of subsequent image processing algorithms. [0003] Image quality evaluation is divided into reference, semi-reference and no-reference methods. Considering that it is impossible to obtain undegraded clear images in practical applications, the image blur evaluation method without reference is more practical. No-r...

AI Technical Summary

Problems solved by technology

However, these methods do not take advantage of the consistency of image information on different scales, and many methods also need to train regression functions, which brings some inconvenience to practical applications.

Images