An Adaptive Compression Sampling Imaging Method Based on Haar Wavelet Sibling Coefficients

Abstract

The invention relates to a self-adaptive compressed sampling imaging method based on a Haar wavelet brother coefficient. The method comprises the following steps: firstly acquiring all wavelet coefficients under a target standard resolution by using a digital micromirror device (DMD); then from the standard resolution, predicting the next layer non-sampled important wavelet coefficient position from the layer sampled coefficient by combining a predetermined wavelet important threshold value according to the relationship between a Haar wavelet father and son coefficient and a brother coefficient, and configuring a wavelet base by using the DMD, and sampling the important wavelet coefficients to obtain a wavelet coefficient matrix with relatively high resolution; repeating predicting and sampling processes till the predetermined imaging resolution is reached to stop sampling, and carrying out wavelet inverse transform to obtain a target scene image. By combining the Haar wavelet brother coefficient information, the predicting precision of the important wavelet coefficient position is improved, the number of times of measurement required by DMD sampling is reduced, and the imaging quality and the imaging speed are improved.

Description

technical field [0001] The invention relates to an imaging technology suitable for a single-arm intensity correlation imaging laser radar system, in particular to an adaptive compression sampling imaging method based on an extended wavelet tree. Background technique [0002] Intensity-correlated imaging, also known as ghost imaging, is a new imaging method developed in recent years ([1] D.V. ‖ Phys. Rev. Lett. 74, 3600–3603 (1995).). In this method, the light emitted by the light source is divided into two paths: one is the signal light path, which illuminates the target, and uses a single-pixel barrel detector to record the light intensity of the target's reflected echo; the other is the reference light path, which uses a high-resolution area array to detect The device obtains the two-dimensional intensity distribution information of the light source. After multiple measurements, the image information of the target can be obtained by measuring the intensity correlation fu...

AI Technical Summary

Problems solved by technology

In response to this problem, some researchers have developed a deterministic matrix that can be generated by hardware [4] ([4] E.Cand`es and J.Romberg, Robust signal recovery from incomplete observations, in Proceedings of the IEEE International Conference on Image Processing, 2006 ,1281–1284.), but these matrices are often designed for certain imaging applications and are not universal

On the other hand, the compressive sensing reconstruction algorithm has high computational complexity. With the increase of image resolution, image reconstruction requires huge computing resources and takes tens of minutes or even longer, making it difficult to apply in practice.

In addition, most reconstruction algorithms require iteration, and the number of iterations is determined by parameters such as residual tolerance, and these parameters are often empirical values, which are only applicable to a certain type of signal, and it is difficult to be universal

Images