Method for fast carrying out edge extraction on fuzzy object

Abstract

The invention discloses a method for fast carrying out edge extraction on a fuzzy object. According to the method, after being modulated by a spatial light modulator (SLM), a light source is projected on the target; after the light source is reflected / transmitted by the target, a detector receives fuzzy image information under the action of a scattering medium; a projected random speckle is moved in four directions according to Roberts and gradient operators, the detector records four data and calculates the partial derivatives of the image in each direction; after a plurality of times of sampling is performed, second-order correlation calculation is performed, and the strong correlation information of the four partial derivatives is obtained, and gradient information is obtained; and two pieces of gradient information are added, so that rapid reconstruction of the information of edges which incline to any directions can be realized. With the method of the invention adopted, the quantity of reconstruction information can be decreased, the difficulty of the recognition of the target can be decreased, and the computation speed of a system can be improved; the fuzzy image can be accurately recognized; and the number of times of sampling required by the reconstruction of a target image is small, and the quantity of data required to be processed by the system is decreased, and hardware implementation can be easier.

Description

technical field [0001] The invention belongs to the field of image processing edge extraction, in particular to a method for rapidly extracting edges from fuzzy objects. Background technique [0002] With the increasingly serious environmental pollution, the smog weather in the city has an impact on people's travel and life. Although the traditional imaging technology is mature at present, there is no fast and effective algorithm for image deblurring. Due to its anti-atmospheric disturbance characteristics, ghost imaging has attracted extensive attention from scholars at home and abroad. Ghost imaging has fast calculation speed and strong anti-scattering ability, but due to its high sampling rate and low signal-to-noise ratio, it has become a key factor restricting the development of ghost imaging. [0003] In recent years, research teams at home and abroad have mainly carried out research on resisting atmospheric turbulence, reducing image noise, and improving the image q...

AI Technical Summary

Problems solved by technology

Although these algorithms have improved the image quality to a certain extent, they have increased the computational complexity of the algorithm, and the Nyquist sampling rate is far greater than 100%, and the image reconstruction speed has not improved.

In 2015, Xue-feng Liu et al. proposed gradient operator ghost imaging (28Dec 2015, Vol.23, No.26, DOI: 10.1364 / OE.23.033802, OPTICSEXPRESS 33803) to quickly reconstruct the edge of the target and reduce the cost of reconstructed images. The required sampling rate, but the edge profile of the target selected by this method has a specific angle, and it is impossible to extract the edge inclined in any direction

In 2016, Tianyi Mao proposed displacement speckle ghost imaging (Volume 8, Number 4, August2016, DOI: 10.1109 / JPHOT.2016.2578934), which can extract contour information at any angle, but the algorithm is complex and the sampling rate is greatly improved compared to GGI. And increase the system data storage burden

Images