[go: up one dir, main page]

CN116132816A - Fourier Panoramic Ghost Imaging Method Based on Logarithmic Linear Transform - Google Patents

Fourier Panoramic Ghost Imaging Method Based on Logarithmic Linear Transform Download PDF

Info

Publication number
CN116132816A
CN116132816A CN202211639498.7A CN202211639498A CN116132816A CN 116132816 A CN116132816 A CN 116132816A CN 202211639498 A CN202211639498 A CN 202211639498A CN 116132816 A CN116132816 A CN 116132816A
Authority
CN
China
Prior art keywords
fourier
panoramic
ghost imaging
radius
spectrum
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202211639498.7A
Other languages
Chinese (zh)
Inventor
曹杰
崔焕�
郝群
周栋
王超
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Institute of Technology BIT
Original Assignee
Beijing Institute of Technology BIT
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Institute of Technology BIT filed Critical Beijing Institute of Technology BIT
Priority to CN202211639498.7A priority Critical patent/CN116132816A/en
Publication of CN116132816A publication Critical patent/CN116132816A/en
Pending legal-status Critical Current

Links

Images

Landscapes

  • Image Processing (AREA)

Abstract

本发明公开了一种基于对数直线变换的傅里叶全景鬼成像方法,包括:步骤一、基于对数直线变换,得到非均匀采样策略;步骤二、基于所述非均匀采样策略构建环形傅里叶基底图案序列;步骤三、基于所述环形傅里叶基底图案序列重构全景鬼成像图像。本发明通过设计环形傅里叶基底图案,利用相移法进行差分处理并且进行傅里叶逆变换来快速重构目标,提升了全景鬼成像的重构速度。

Figure 202211639498

The invention discloses a Fourier panoramic ghost imaging method based on logarithmic linear transformation, comprising: step 1, obtaining a non-uniform sampling strategy based on logarithmic linear transformation; step 2, constructing a ring Fu Fourier basis pattern sequence; step 3, reconstructing the panoramic ghost imaging image based on the annular Fourier basis pattern sequence. The present invention designs a ring-shaped Fourier base pattern, uses a phase shift method to perform differential processing and performs Fourier inverse transformation to quickly reconstruct a target, thereby improving the reconstruction speed of panoramic ghost imaging.

Figure 202211639498

Description

基于对数直线变换的傅里叶全景鬼成像方法Fourier Panoramic Ghost Imaging Method Based on Logarithmic Linear Transform

技术领域technical field

本发明属于光电成像技术领域,具体涉及一种基于对数直线变换的傅里叶全景鬼成像方法。The invention belongs to the technical field of photoelectric imaging, and in particular relates to a Fourier panoramic ghost imaging method based on logarithmic linear transformation.

背景技术Background technique

全景鬼成像是一种实现周向360°大视场的新型关联成像方法,其利用曲面镜将用于调制光源的环形散斑图案反射至周围360°的目标场景,随后用一个桶探测器收集反射回来的光强涨落情况,并与相应的基于对数极坐标变换映射的矩形散斑序列进行关联计算,便可以直接重构无畸变展开的全景目标场景。相比于传统全景成像方法,全景鬼成像具有抗干扰能力强、成像分辨率超越衍射极限、可在不同波段成像、探测灵敏度高等成像优势,可用于解决传统全景成像技术遇到的问题与挑战,来扩展全景成像的应用需求。但基于传统二阶关联特性进行全景目标重构的理论,现有全景鬼成像需要大量采样次数和重构时间,无法实现快速高质量成像,难以兼顾大视场、高质量和实时性,阻碍了全景鬼成像的实用化进程。Panoramic ghost imaging is a new correlative imaging method to achieve a large circumferential 360° field of view. It uses a curved mirror to reflect the annular speckle pattern used to modulate the light source to the surrounding 360° target scene, and then collects it with a barrel detector. The reflected light intensity fluctuations are correlated with the corresponding rectangular speckle sequence based on the logarithmic polar coordinate transformation mapping, and the panoramic target scene without distortion can be directly reconstructed. Compared with traditional panoramic imaging methods, panoramic ghost imaging has the advantages of strong anti-interference ability, imaging resolution beyond the diffraction limit, imaging in different bands, and high detection sensitivity. It can be used to solve the problems and challenges encountered by traditional panoramic imaging technologies. To expand the application requirements of panoramic imaging. However, based on the theory of panoramic target reconstruction based on traditional second-order correlation characteristics, the existing panoramic ghost imaging requires a large number of sampling times and reconstruction time, which cannot achieve fast and high-quality imaging, and it is difficult to take into account large field of view, high quality and real-time performance, which hinders The practical progress of panoramic ghost imaging.

发明内容Contents of the invention

为了实现快速高质量的全景鬼成像,解决鬼成像大视场、高质量和实时性难以兼顾的问题,本发明提出一种基于对数直线变换实现非均匀采样策略的傅里叶全景鬼成像方法,将傅里叶单像素成像的概念引入全景鬼成像理论中,提出傅里叶全景鬼成像方法,通过具有相移的环形傅里叶基底图案来获取目标的傅里叶频谱信息,随后直接进行快速傅里叶逆变换就可以完成成像;实现低采样率下的高质量傅里叶全景鬼成像,进一步降低采样时间来提升傅里叶全景鬼成像的实时性。In order to achieve fast and high-quality panoramic ghost imaging and solve the problem of large field of view, high quality and real-time performance of ghost imaging, this invention proposes a Fourier panoramic ghost imaging method based on logarithmic linear transformation to achieve non-uniform sampling strategy , the concept of Fourier single-pixel imaging is introduced into the theory of panoramic ghost imaging, and a Fourier panoramic ghost imaging method is proposed, which obtains the Fourier spectrum information of the target through the annular Fourier base pattern with phase shift, and then directly performs Fast Fourier inverse transform can complete the imaging; realize high-quality Fourier panoramic ghost imaging at low sampling rate, and further reduce the sampling time to improve the real-time performance of Fourier panoramic ghost imaging.

为实现上述目的,本发明提供了如下方案:To achieve the above object, the present invention provides the following scheme:

一种基于对数直线变换的傅里叶全景鬼成像方法,包括:A Fourier panoramic ghost imaging method based on logarithmic linear transformation, comprising:

步骤一、基于对数直线变换,得到非均匀采样策略;Step 1. Obtain a non-uniform sampling strategy based on logarithmic linear transformation;

步骤二、基于所述非均匀采样策略构建环形傅里叶基底图案序列;Step 2, constructing a circular Fourier base pattern sequence based on the non-uniform sampling strategy;

步骤三、基于所述环形傅里叶基底图案序列重构全景鬼成像图像。Step 3. Reconstructing a panoramic ghost imaging image based on the circular Fourier basis pattern sequence.

优选地,得到所述非均匀采样策略的方法包括:Preferably, the method for obtaining the non-uniform sampling strategy includes:

初始化傅里叶全景鬼成像参数;Initialize Fourier panoramic ghost imaging parameters;

基于所述傅里叶全景鬼成像参数和所述对数直线变换,设计非均匀采样策略。Based on the Fourier panoramic ghost imaging parameters and the logarithmic linear transformation, a non-uniform sampling strategy is designed.

优选地,所述非均匀采样策略包括:Preferably, the non-uniform sampling strategy includes:

Figure BDA0004006440010000021
Figure BDA0004006440010000021

式中,

Figure BDA0004006440010000022
表示p-q坐标系中以低频中心点向p方向扩展的第u层选择间隔的半径,
Figure BDA0004006440010000023
表示以低频中心点向q方向扩展的第v层选择间隔的半径;p(u)和q(v)分别表示p方向第u个选择的频谱坐标和q方向第v个选择的频谱坐标;m0表示低频区域的p方向半径,n0表示低频区域的q方向半径,U表示频谱在p方向上的采样点数,V表示频谱在q方向上的采样点数,(fp0,fq0)表示全景傅里叶频谱的中心点坐标,α1表示采样的频谱系数在p向间隔增长因子,α2表示采样的频谱系数在q向间隔增长因子。In the formula,
Figure BDA0004006440010000022
Indicates the radius of the u-th layer selection interval extending from the low-frequency center point to the p direction in the pq coordinate system,
Figure BDA0004006440010000023
Indicates the radius of the selection interval of the vth layer extending from the low-frequency center point to the q direction; p(u) and q(v) represent the spectrum coordinates of the uth selection in the p direction and the spectrum coordinates of the vth selection in the q direction, respectively; m0 Indicates the radius of the p-direction of the low-frequency region, n0 represents the radius of the q-direction of the low-frequency region, U represents the number of sampling points of the spectrum in the direction of p, V represents the number of sampling points of the spectrum in the direction of q, (f p0 , f q0 ) represents the panoramic Fourier The coordinates of the center point of the leaf spectrum, α 1 represents the growth factor of the sampled spectral coefficients in the p-direction interval, and α 2 represents the growth factor of the sampled spectral coefficients in the q-direction interval.

优选地,构建所述环形傅里叶基底图案序列的方法包括:Preferably, the method for constructing the circular Fourier base pattern sequence comprises:

基于所述非均匀采样策略设计全景傅里叶基底图案模型;Designing a panoramic Fourier base pattern model based on the non-uniform sampling strategy;

基于所述全景傅里叶基底图案模型,构建所述环形傅里叶基底图案序列。The circular Fourier basis pattern sequence is constructed based on the panoramic Fourier basis pattern model.

优选地,所述全景傅里叶基底图案模型包括:Preferably, the panoramic Fourier base pattern model includes:

Puv=a+b×cos(2πfp(u)p(u)+2πfq(v)q(v)+φ),P uv =a+b×cos(2πf p(u) p(u)+2πf q(v) q(v)+φ),

式中,(p(u),q(v))表示非均匀采样策略选择的频谱系数对应的空间坐标;(fp(u),fq(v))表示非均匀采样策略选择的频谱系数;φ表示初始相位;Puv表示在相应的频谱系数下所对应的全景傅里叶基底图案模型;a表示平均光强,b表示对比度。In the formula, (p(u), q(v)) represents the spatial coordinates corresponding to the spectral coefficients selected by the non-uniform sampling strategy; (f p(u) , f q(v) ) represents the spectral coefficients selected by the non-uniform sampling strategy ; φ represents the initial phase; P uv represents the corresponding panoramic Fourier basis pattern model under the corresponding spectral coefficient; a represents the average light intensity, and b represents the contrast.

优选地,所述环形傅里叶基底图案序列包括:Preferably, the sequence of circular Fourier basis patterns comprises:

Figure BDA0004006440010000041
Figure BDA0004006440010000041

式中,rp表示环形傅里叶散斑结构第p环的半径,θq表示第q个像素所对应的角度,Q表示每环平均划分的单元数,ε表示环间增长系数,r0表示中央凹盲孔区域的外环半径,rc1表示第1环的中心半径,rcp表示第p环的中心半径,ξp表示第p环在对数极坐标系下对应的数值。In the formula, r p represents the radius of the pth ring of the annular Fourier speckle structure, θ q represents the angle corresponding to the qth pixel, Q represents the average number of units divided by each ring, ε represents the inter-ring growth coefficient, and r 0 Indicates the outer ring radius of the central concave blind hole area, rc 1 indicates the center radius of the first ring, rc p indicates the center radius of the p-th ring, and ξ p indicates the corresponding value of the p-th ring in the logarithmic polar coordinate system.

优选地,重构所述全景鬼成像图像的方法包括:Preferably, the method for reconstructing the panoramic ghost imaging image includes:

基于所述环形傅里叶基底图案序列调制光源,用调制后的所述光源照射在曲面反射镜上;Modulating a light source based on the annular Fourier base pattern sequence, and irradiating the curved mirror with the modulated light source;

所述光源经曲面反射镜反射,得到反射光,所述反射光照明周围360°视场的目标场景;The light source is reflected by the curved mirror to obtain reflected light, and the reflected light illuminates the target scene in the surrounding 360° field of view;

照明周围360°视场的目标场景后的所述反射光发生漫反射,采集漫反射后带有目标信息的光强;Diffuse reflection occurs in the reflected light after illuminating the target scene of the surrounding 360 ° field of view, and the light intensity with target information after the diffuse reflection is collected;

将不同初始相位的所述环形傅里叶基底图案序列对应的所述光强进行差分处理,得到傅里叶频谱;performing differential processing on the light intensities corresponding to the annular Fourier base pattern sequences with different initial phases to obtain a Fourier spectrum;

对所述傅里叶频谱进行傅里叶逆变换,得到所述全景鬼成像图像。Inverse Fourier transform is performed on the Fourier spectrum to obtain the panoramic ghost imaging image.

优选地,所述光强的计算方法包括:Preferably, the calculation method of the light intensity includes:

Figure BDA0004006440010000051
Figure BDA0004006440010000051

式中,Ti表示利用第i张初始相位为φj的环形傅里叶基底图案Si调制光源得到的光强值,o(x,y)表示照射到的周围360°全向环形目标。In the formula, T i represents the light intensity value obtained by modulating the light source with the i-th annular Fourier base pattern S i whose initial phase is φ j , and o(x, y) represents the irradiated surrounding 360° omnidirectional annular target.

本发明的有益效果为:The beneficial effects of the present invention are:

(1)本发明公开的一种基于对数直线变换的傅里叶全景鬼成像方法,通过设计环形傅里叶基底图案,利用相移法进行差分处理并且进行傅里叶逆变换来快速重构目标,提升了全景鬼成像的重构速度。(1) A Fourier panoramic ghost imaging method based on logarithmic linear transformation disclosed in the present invention, by designing a circular Fourier base pattern, using the phase shift method to perform differential processing and performing inverse Fourier transform to quickly reconstruct The goal is to improve the reconstruction speed of panoramic ghost imaging.

(2)本发明公开的一种基于对数直线变换的傅里叶全景鬼成像方法,可基于对数直线变换自由灵活地对全景傅里叶频谱进行非均匀采样,提高欠采样率下的高效频谱采样,能够降低全景鬼成像的采样次数,降低了全景鬼成像的采样时间。(2) A Fourier panoramic ghost imaging method based on logarithmic linear transformation disclosed in the present invention can freely and flexibly perform non-uniform sampling on the panoramic Fourier spectrum based on logarithmic linear transformation, and improve the efficiency under the under-sampling rate. Spectrum sampling can reduce the sampling frequency of panoramic ghost imaging and reduce the sampling time of panoramic ghost imaging.

(3)本发明公开的一种基于对数直线变换的傅里叶全景鬼成像方法,能够大幅度提升全景鬼成像的成像时间,提高全景鬼成像的实时性,为同时兼顾大视场、高分辨、实时性的鬼成像提供了一条有效的途径,进而满足更多大视场应用领域的成像需求,加速了鬼成像的实用化进程。(3) A Fourier panoramic ghost imaging method based on logarithmic linear transformation disclosed in the present invention can greatly increase the imaging time of panoramic ghost imaging, improve the real-time performance of panoramic ghost imaging, and take into account both large field of view and high The resolution and real-time ghost imaging provides an effective way to meet the imaging needs of more large field of view applications and accelerate the practical process of ghost imaging.

附图说明Description of drawings

为了更清楚地说明本发明的技术方案,下面对实施例中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动性的前提下,还可以根据这些附图获得其他的附图。In order to illustrate the technical solution of the present invention more clearly, the accompanying drawings that need to be used in the embodiments are briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. Technical personnel can also obtain other drawings based on these drawings without paying creative labor.

图1为本发明实施例一方法流程示意图;Fig. 1 is a schematic flow chart of a method according to Embodiment 1 of the present invention;

图2为本发明实施例一环形傅里叶基底图案结构原理图;图2(a)为环形傅里叶基底图案结构成像区域的细节放大图,图2(b)为环形傅里叶基底图案结构图,图2(c)为一个环形傅里叶散斑图案示例图;Fig. 2 is a principle diagram of the structure of the annular Fourier base pattern according to the embodiment of the present invention; Fig. 2 (a) is a detailed enlarged view of the imaging area of the annular Fourier base pattern structure, and Fig. 2 (b) is the annular Fourier base pattern Structural diagram, Figure 2(c) is an example diagram of a circular Fourier speckle pattern;

图3为本发明实施例一基于对数直线变换的全景傅里叶频谱的非均匀采样频谱图。FIG. 3 is a non-uniform sampling spectrum diagram of a panoramic Fourier spectrum based on logarithmic linear transformation according to Embodiment 1 of the present invention.

具体实施方式Detailed ways

下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

为使本发明的上述目的、特征和优点能够更加明显易懂,下面结合附图和具体实施方式对本发明作进一步详细的说明。In order to make the above objects, features and advantages of the present invention more comprehensible, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

实施例一Embodiment one

如图1所示,为本发明基于对数直线变换的傅里叶全景鬼成像方法流程示意图,包括以下步骤:As shown in Figure 1, it is a schematic flow chart of the Fourier panoramic ghost imaging method based on logarithmic linear transformation in the present invention, including the following steps:

步骤一、基于对数直线变换,得到非均匀采样策略;Step 1. Obtain a non-uniform sampling strategy based on logarithmic linear transformation;

具体包括:Specifically include:

(1)初始化傅里叶全景鬼成像参数;(1) Initialize Fourier panoramic ghost imaging parameters;

设傅里叶全景鬼成像图像的空间分辨率为P×Q,频域分辨率同样设置为P×Q,并对频率系数进行归一化离散处理,处理方法如下:Assume that the spatial resolution of the Fourier panoramic ghost imaging image is P×Q, and the frequency domain resolution is also set as P×Q, and the frequency coefficients are normalized and discretely processed. The processing method is as follows:

Figure BDA0004006440010000071
Figure BDA0004006440010000071

式中,fp代表p方向第p个频率系数,fq代表q方向第q个频率系数。In the formula, f p represents the pth frequency coefficient in the p direction, and f q represents the qth frequency coefficient in the q direction.

本实施例中,设傅里叶全景鬼成像图像的空间分辨率为20×102,频域分辨率同样设置为20×102。In this embodiment, the spatial resolution of the Fourier panoramic ghost imaging image is set to 20×10 2 , and the frequency domain resolution is also set to 20×10 2 .

(2)基于傅里叶全景鬼成像参数和对数直线变换,设计非均匀采样策略。(2) Based on Fourier panoramic ghost imaging parameters and logarithmic linear transformation, a non-uniform sampling strategy is designed.

设傅里叶频谱在p方向采样率为sp,在q方向采样率为sq,则整个采样过程需要U×V×2次,其中U=P×sp,V=Q×sq。本实施例中,p方向采样率和q方向采样率均设为0.3,则整个采样过程需要6×31×2次。如图3所示,为实现高效采样,根据频谱图的高低分布,构建基于对数直线变换的非均匀采样策略:Assuming that the sampling rate of the Fourier spectrum in the p direction is s p , and the sampling rate in the q direction is s q , the whole sampling process needs U×V×2 times, where U=P×s p , V=Q×s q . In this embodiment, the sampling rate in the p-direction and the sampling rate in the q-direction are both set to 0.3, and the entire sampling process requires 6×31×2 times. As shown in Figure 3, in order to achieve efficient sampling, according to the high and low distribution of the spectrogram, a non-uniform sampling strategy based on logarithmic linear transformation is constructed:

在低频区域进行全采;在高频区域,对频谱系数以指数形式的间隔分别向p、q两个方向进行选择采样。设全景傅里叶频谱的中心点坐标为(fp0,fq0),设为选择进行全采的低频区域的p方向半径为m0,q向半径为n0;高频区域中,选择采样的频谱系数在p向间隔增长因子为α1,q向间隔增长因子为α2。本实施例中,全景傅里叶频谱的中心点坐标设为(11,52),选择进行全采的低频区域的p方向半径为2,q向半径为8,在高频区域中,选择采样的频谱系数在p向间隔增长因子为2.2361,q向间隔增长因子为1.2605。Full sampling is carried out in the low-frequency region; in the high-frequency region, spectral coefficients are selectively sampled in two directions p and q at exponential intervals. Set the coordinates of the center point of the panoramic Fourier spectrum as (f p0 , f q0 ), set the radius in the p direction of the low-frequency region selected for full sampling as m0, and the radius in the q direction as n0; in the high-frequency region, select the sampled spectrum The increase factor of the coefficient in the p-direction interval is α 1 , and the increase factor in the q-direction interval is α 2 . In this embodiment, the coordinates of the central point of the panoramic Fourier spectrum are set to (11, 52), and the radius in the p direction of the low-frequency area selected for full sampling is 2, and the radius in the q direction is 8. In the high-frequency area, the sampling The spectral coefficient of the p-direction interval growth factor is 2.2361, and the q-direction interval growth factor is 1.2605.

则基于对数直线变换的非均匀采样策略模型如下:Then the non-uniform sampling strategy model based on logarithmic linear transformation is as follows:

Figure BDA0004006440010000081
Figure BDA0004006440010000081

式中:

Figure BDA0004006440010000082
表示p-q坐标系中以低频中心点向p方向扩展的第u层选择间隔的半径,
Figure BDA0004006440010000083
表示以低频中心点向q方向扩展的第v层选择间隔的半径;p(u)和q(v)分别表示p方向第u个选择的频谱坐标和q方向第v个选择的频谱坐标;m0表示低频区域的p方向半径,n0表示低频区域的q方向半径,U表示频谱在p方向上的采样点数,V表示频谱在q方向上的采样点数,(fp0,fq0)表示全景傅里叶频谱的中心点坐标,α1表示采样的频谱系数在p向间隔增长因子,α2表示采样的频谱系数在q向间隔增长因子。In the formula:
Figure BDA0004006440010000082
Indicates the radius of the u-th layer selection interval extending from the low-frequency center point to the p direction in the pq coordinate system,
Figure BDA0004006440010000083
Indicates the radius of the selection interval of the vth layer extending from the low-frequency center point to the q direction; p(u) and q(v) represent the spectrum coordinates of the uth selection in the p direction and the spectrum coordinates of the vth selection in the q direction, respectively; m0 Indicates the radius of the p-direction of the low-frequency region, n0 represents the radius of the q-direction of the low-frequency region, U represents the number of sampling points of the spectrum in the direction of p, V represents the number of sampling points of the spectrum in the direction of q, (f p0 , f q0 ) represents the panoramic Fourier The coordinates of the center point of the leaf spectrum, α 1 represents the growth factor of the sampled spectral coefficients in the p-direction interval, and α 2 represents the growth factor of the sampled spectral coefficients in the q-direction interval.

步骤二、基于非均匀采样策略构建环形傅里叶基底图案序列;Step 2, constructing a circular Fourier base pattern sequence based on a non-uniform sampling strategy;

具体包括:Specifically include:

(1)基于非均匀采样策略设计全景傅里叶基底图案模型;(1) Design a panoramic Fourier base pattern model based on a non-uniform sampling strategy;

根据非均匀采样策略选择的频谱系数设计全景傅里叶基底图案模型,计算方法如下:According to the spectral coefficients selected by the non-uniform sampling strategy, the panoramic Fourier basis pattern model is designed, and the calculation method is as follows:

Puv=1+0.5×cos(2πfp(u)p(u)+2πfq(v)q(v)+φ),P uv =1+0.5×cos(2πf p(u) p(u)+2πf q(v) q(v)+φ),

式中,(p(u),q(v))表示非均匀采样策略选择的频谱系数对应的空间坐标;(fp(u),fq(v))表示非均匀采样策略选择的频谱系数;φ表示初始相位,根据四步相移法可取值为:0,π/2,π,3π/2;Puv表示在相应的频谱系数下所对应的全景傅里叶基底图案模型;a表示平均光强,b表示对比度。In the formula, (p(u), q(v)) represents the spatial coordinates corresponding to the spectral coefficients selected by the non-uniform sampling strategy; (f p(u) , f q(v) ) represents the spectral coefficients selected by the non-uniform sampling strategy ; φ represents the initial phase, according to the four-step phase shift method, the possible values are: 0, π/2, π, 3π/2; P uv represents the corresponding panoramic Fourier basis pattern model under the corresponding spectral coefficient; a Indicates the average light intensity, and b indicates the contrast.

(2)基于全景傅里叶基底图案模型,构建环形傅里叶基底图案序列。(2) Based on the panoramic Fourier basis pattern model, a circular Fourier basis pattern sequence is constructed.

如图2所示,环形傅里叶基底图案包括中央凹盲孔区域和成像区域,中央凹盲孔区域的外环半径r0内为盲区,r0外为成像区域。根据全景鬼成像图像的分辨率P×Q,成像区域划分为P环,每环均匀分布Q个像素,并且用ε表示环间增长系数。本实施例中,根据全景鬼成像图像的分辨率20×102,成像区域划分为20环,每环均匀分布102个像素。As shown in Fig. 2, the annular Fourier base pattern includes the central fovea blind hole area and the imaging area, the outer ring radius r 0 of the central fovea blind hole area is the blind area, and the outside r 0 is the imaging area. According to the resolution P×Q of the panoramic ghost imaging image, the imaging area is divided into P rings, and Q pixels are uniformly distributed in each ring, and ε represents the growth coefficient between rings. In this embodiment, according to the resolution of the panoramic ghost imaging image of 20×102, the imaging area is divided into 20 rings, and 102 pixels are evenly distributed in each ring.

环形傅里叶基底图案序列包括:The sequence of circular Fourier basis patterns includes:

Figure BDA0004006440010000091
Figure BDA0004006440010000091

式中,rp表示环形傅里叶散斑结构第p环的半径,θq表示第q个像素所对应的角度,Q表示每环平均划分的单元数,ε表示环间增长系数,r0表示中央凹盲孔区域的外环半径,rc1表示第1环的中心半径,rcp表示第p环的中心半径,ξp表示第p环在对数极坐标系下对应的数值。In the formula, r p represents the radius of the pth ring of the annular Fourier speckle structure, θ q represents the angle corresponding to the qth pixel, Q represents the average number of units divided by each ring, ε represents the inter-ring growth coefficient, and r 0 Indicates the outer ring radius of the central concave blind hole area, rc 1 indicates the center radius of the first ring, rc p indicates the center radius of the p-th ring, and ξ p indicates the corresponding value of the p-th ring in the logarithmic polar coordinate system.

步骤三、基于环形傅里叶基底图案序列重构全景鬼成像图像。Step 3, reconstructing the panoramic ghost imaging image based on the circular Fourier basis pattern sequence.

具体包括:Specifically include:

(1)基于环形傅里叶基底图案序列调制光源,用调制后的光源照射在曲面反射镜上;(1) Modulate the light source based on the circular Fourier base pattern sequence, and irradiate the curved mirror with the modulated light source;

(2)光源经曲面反射镜反射,得到反射光,反射光照明周围360°视场的目标场景;(2) The light source is reflected by the curved mirror to obtain reflected light, which illuminates the target scene in the surrounding 360° field of view;

(3)照明周围360°视场的目标场景后的反射光发生漫反射,采集漫反射后带有目标信息的光强;(3) The reflected light after illuminating the target scene in the surrounding 360° field of view undergoes diffuse reflection, and the light intensity with target information after the diffuse reflection is collected;

光强的计算方法包括:The calculation methods of light intensity include:

Figure BDA0004006440010000101
Figure BDA0004006440010000101

式中,Ti表示利用第i张初始相位为φj的环形傅里叶基底图案Si调制光源得到的光强值,o(x,y)表示照射到的周围360°全向环形目标。In the formula, T i represents the light intensity value obtained by modulating the light source with the i-th annular Fourier base pattern S i whose initial phase is φ j , and o(x, y) represents the irradiated surrounding 360° omnidirectional annular target.

(4)将不同初始相位的环形傅里叶基底图案序列对应的光强进行差分处理,得到傅里叶频谱;计算方法如下:(4) Differentially process the light intensities corresponding to the annular Fourier base pattern sequences of different initial phases to obtain the Fourier spectrum; the calculation method is as follows:

首先将U×V对具有不同初始相位的环形傅里叶基底图案所对应的光强值进行差分处理来得到相应坐标下的频谱系数值:First, U×V performs differential processing on the light intensity values corresponding to the annular Fourier base patterns with different initial phases to obtain the spectral coefficient values at the corresponding coordinates:

Figure BDA0004006440010000102
Figure BDA0004006440010000102

式中,

Figure BDA0004006440010000103
表示透射初始相位为0、频谱系数坐标为(p(u),q(v))的环形傅里叶基底图案所对应的光强值,
Figure BDA0004006440010000111
表示透射初始相位为π/2、频谱系数坐标为(p(u),q(v))的环形傅里叶基底图案所对应的光强值,
Figure BDA0004006440010000112
表示透射初始相位为π、频谱系数坐标为(p(u),q(v))的环形傅里叶基底图案所对应的光强值,
Figure BDA0004006440010000113
表示透射初始相位为3π/2、频谱系数坐标为(p(u),q(v))的环形傅里叶基底图案所对应的光强值,fp(u),q(v)表示坐标为(p(u),q(v))的频谱系数,j表示虚部。In the formula,
Figure BDA0004006440010000103
Indicates the light intensity value corresponding to the circular Fourier basis pattern with the transmission initial phase of 0 and spectral coefficient coordinates of (p(u),q(v)),
Figure BDA0004006440010000111
Indicates the light intensity value corresponding to the circular Fourier basis pattern with the transmission initial phase of π/2 and the spectral coefficient coordinates of (p(u),q(v)),
Figure BDA0004006440010000112
Indicates the light intensity value corresponding to the circular Fourier basis pattern with the transmission initial phase as π and spectral coefficient coordinates as (p(u),q(v)),
Figure BDA0004006440010000113
Indicates the light intensity value corresponding to the annular Fourier basis pattern with the transmission initial phase of 3π/2 and the spectral coefficient coordinates of (p(u),q(v)), f p(u),q(v) represent the coordinates is the spectral coefficient of (p(u),q(v)), and j represents the imaginary part.

(5)对傅里叶频谱进行傅里叶逆变换,得到全景鬼成像图像。(5) Inverse Fourier transform is performed on the Fourier spectrum to obtain a panoramic ghost imaging image.

计算方法如下:The calculation method is as follows:

o′(p,q)=F-1(fpq)o'(p,q)=F -1 (f pq )

式中,o'表示重构的全景鬼成像图像,F-1表示傅里叶逆变换符号,fpq表示P×Q全景图像的傅里叶频谱系数矩阵。In the formula, o' represents the reconstructed panoramic ghost imaging image, F -1 represents the inverse Fourier transform symbol, and f pq represents the Fourier spectral coefficient matrix of the P×Q panoramic image.

以上所述的实施例仅是对本发明优选方式进行的描述,并非对本发明的范围进行限定,在不脱离本发明设计精神的前提下,本领域普通技术人员对本发明的技术方案做出的各种变形和改进,均应落入本发明权利要求书确定的保护范围内。The above-mentioned embodiments are only descriptions of the preferred modes of the present invention, and do not limit the scope of the present invention. Variations and improvements should fall within the scope of protection defined by the claims of the present invention.

Claims (8)

1.一种基于对数直线变换的傅里叶全景鬼成像方法,其特征在于,包括:1. A Fourier panorama ghost imaging method based on logarithmic rectilinear transform, is characterized in that, comprises: 步骤一、基于对数直线变换,得到非均匀采样策略;Step 1. Obtain a non-uniform sampling strategy based on logarithmic linear transformation; 步骤二、基于所述非均匀采样策略构建环形傅里叶基底图案序列;Step 2, constructing a circular Fourier base pattern sequence based on the non-uniform sampling strategy; 步骤三、基于所述环形傅里叶基底图案序列重构全景鬼成像图像。Step 3. Reconstructing a panoramic ghost imaging image based on the circular Fourier basis pattern sequence. 2.根据权利要求1所述基于对数直线变换的傅里叶全景鬼成像方法,其特征在于,得到所述非均匀采样策略的方法包括:2. according to the described Fourier panoramic ghost imaging method based on logarithmic linear transform according to claim 1, it is characterized in that, the method that obtains described non-uniform sampling strategy comprises: 初始化傅里叶全景鬼成像参数;Initialize Fourier panoramic ghost imaging parameters; 基于所述傅里叶全景鬼成像参数和所述对数直线变换,设计非均匀采样策略。Based on the Fourier panoramic ghost imaging parameters and the logarithmic linear transformation, a non-uniform sampling strategy is designed. 3.根据权利要求2所述基于对数直线变换的傅里叶全景鬼成像方法,其特征在于,所述非均匀采样策略包括:3. the Fourier panoramic ghost imaging method based on logarithmic linear transform according to claim 2, is characterized in that, described non-uniform sampling strategy comprises:
Figure FDA0004006439000000011
Figure FDA0004006439000000011
式中,
Figure FDA0004006439000000012
表示p-q坐标系中以低频中心点向p方向扩展的第u层选择间隔的半径,
Figure FDA0004006439000000013
表示以低频中心点向q方向扩展的第v层选择间隔的半径;p(u)和q(v)分别表示p方向第u个选择的频谱坐标和q方向第v个选择的频谱坐标;m0表示低频区域的p方向半径,n0表示低频区域的q方向半径,U表示频谱在p方向上的采样点数,V表示频谱在q方向上的采样点数,(fp0,fq0)表示全景傅里叶频谱的中心点坐标,α1表示采样的频谱系数在p向间隔增长因子,α2表示采样的频谱系数在q向间隔增长因子。
In the formula,
Figure FDA0004006439000000012
Indicates the radius of the u-th layer selection interval extending from the low-frequency center point to the p direction in the pq coordinate system,
Figure FDA0004006439000000013
Indicates the radius of the selection interval of the vth layer extending from the low-frequency center point to the q direction; p(u) and q(v) represent the spectrum coordinates of the uth selection in the p direction and the spectrum coordinates of the vth selection in the q direction, respectively; m0 Indicates the radius of the p-direction of the low-frequency region, n0 represents the radius of the q-direction of the low-frequency region, U represents the number of sampling points of the spectrum in the direction of p, V represents the number of sampling points of the spectrum in the direction of q, (f p0 , f q0 ) represents the panoramic Fourier The coordinates of the center point of the leaf spectrum, α 1 represents the growth factor of the sampled spectral coefficients in the p-direction interval, and α 2 represents the growth factor of the sampled spectral coefficients in the q-direction interval.
4.根据权利要求1所述基于对数直线变换的傅里叶全景鬼成像方法,其特征在于,构建所述环形傅里叶基底图案序列的方法包括:4. according to the described Fourier panoramic ghost imaging method based on logarithmic linear transform according to claim 1, it is characterized in that, the method for constructing described annular Fourier basis pattern sequence comprises: 基于所述非均匀采样策略设计全景傅里叶基底图案模型;Designing a panoramic Fourier base pattern model based on the non-uniform sampling strategy; 基于所述全景傅里叶基底图案模型,构建所述环形傅里叶基底图案序列。The circular Fourier basis pattern sequence is constructed based on the panoramic Fourier basis pattern model. 5.根据权利要求4所述基于对数直线变换的傅里叶全景鬼成像方法,其特征在于,所述全景傅里叶基底图案模型包括:5. the Fourier panoramic ghost imaging method based on logarithmic linear transformation according to claim 4, is characterized in that, described panoramic Fourier base pattern model comprises: Puv=a+b×cos(2πfp(u)p(u)+2πfq(v)q(v)+φ),P uv =a+b×cos(2πf p(u) p(u)+2πf q(v) q(v)+φ), 式中,(p(u),q(v))表示非均匀采样策略选择的频谱系数对应的空间坐标;(fp(u),fq(v))表示非均匀采样策略选择的频谱系数;φ表示初始相位;Puv表示在相应的频谱系数下所对应的全景傅里叶基底图案模型;a表示平均光强,b表示对比度。In the formula, (p(u), q(v)) represents the spatial coordinates corresponding to the spectral coefficients selected by the non-uniform sampling strategy; (f p(u) , f q(v) ) represents the spectral coefficients selected by the non-uniform sampling strategy ; φ represents the initial phase; P uv represents the corresponding panoramic Fourier basis pattern model under the corresponding spectral coefficient; a represents the average light intensity, and b represents the contrast. 6.根据权利要求4所述基于对数直线变换的傅里叶全景鬼成像方法,其特征在于,所述环形傅里叶基底图案序列包括:6. according to the described Fourier panoramic ghost imaging method based on logarithmic straight line transform of claim 4, it is characterized in that, described annular Fourier base pattern sequence comprises:
Figure FDA0004006439000000031
Figure FDA0004006439000000031
式中,rp表示环形傅里叶散斑结构第p环的半径,θq表示第q个像素所对应的角度,Q表示每环平均划分的单元数,ε表示环间增长系数,r0表示中央凹盲孔区域的外环半径,rc1表示第1环的中心半径,rcp表示第p环的中心半径,ξp表示第p环在对数极坐标系下对应的数值。In the formula, r p represents the radius of the pth ring of the annular Fourier speckle structure, θ q represents the angle corresponding to the qth pixel, Q represents the average number of units divided by each ring, ε represents the inter-ring growth coefficient, and r 0 Indicates the outer ring radius of the central concave blind hole area, rc 1 indicates the center radius of the first ring, rc p indicates the center radius of the p-th ring, and ξ p indicates the corresponding value of the p-th ring in the logarithmic polar coordinate system.
7.根据权利要求1所述基于对数直线变换的傅里叶全景鬼成像方法,其特征在于,重构所述全景鬼成像图像的方法包括:7. according to the described Fourier panoramic ghost imaging method based on logarithmic linear transform according to claim 1, it is characterized in that, the method for reconstructing described panoramic ghost imaging image comprises: 基于所述环形傅里叶基底图案序列调制光源,用调制后的所述光源照射在曲面反射镜上;Modulating a light source based on the annular Fourier base pattern sequence, and irradiating the curved mirror with the modulated light source; 所述光源经曲面反射镜反射,得到反射光,所述反射光照明周围360°视场的目标场景;The light source is reflected by the curved mirror to obtain reflected light, and the reflected light illuminates the target scene in the surrounding 360° field of view; 照明周围360°视场的目标场景后的所述反射光发生漫反射,采集漫反射后带有目标信息的光强;Diffuse reflection occurs in the reflected light after illuminating the target scene of the surrounding 360 ° field of view, and the light intensity with target information after the diffuse reflection is collected; 将不同初始相位的所述环形傅里叶基底图案序列对应的所述光强进行差分处理,得到傅里叶频谱;performing differential processing on the light intensities corresponding to the annular Fourier base pattern sequences with different initial phases to obtain a Fourier spectrum; 对所述傅里叶频谱进行傅里叶逆变换,得到所述全景鬼成像图像。Inverse Fourier transform is performed on the Fourier spectrum to obtain the panoramic ghost imaging image. 8.根据权利要求7所述基于对数直线变换的傅里叶全景鬼成像方法,其特征在于,所述光强的计算方法包括:8. the Fourier panoramic ghost imaging method based on logarithmic linear transform according to claim 7, is characterized in that, the computing method of described light intensity comprises:
Figure FDA0004006439000000041
Figure FDA0004006439000000041
式中,Ti表示利用第i张初始相位为φj的环形傅里叶基底图案Si调制光源得到的光强值,o(x,y)表示照射到的周围360°全向环形目标。In the formula, T i represents the light intensity value obtained by modulating the light source with the i-th annular Fourier base pattern S i whose initial phase is φ j , and o(x, y) represents the irradiated surrounding 360° omnidirectional annular target.
CN202211639498.7A 2022-12-19 2022-12-19 Fourier Panoramic Ghost Imaging Method Based on Logarithmic Linear Transform Pending CN116132816A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211639498.7A CN116132816A (en) 2022-12-19 2022-12-19 Fourier Panoramic Ghost Imaging Method Based on Logarithmic Linear Transform

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211639498.7A CN116132816A (en) 2022-12-19 2022-12-19 Fourier Panoramic Ghost Imaging Method Based on Logarithmic Linear Transform

Publications (1)

Publication Number Publication Date
CN116132816A true CN116132816A (en) 2023-05-16

Family

ID=86309247

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211639498.7A Pending CN116132816A (en) 2022-12-19 2022-12-19 Fourier Panoramic Ghost Imaging Method Based on Logarithmic Linear Transform

Country Status (1)

Country Link
CN (1) CN116132816A (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2016101008A1 (en) * 2014-12-23 2016-06-30 Canon Kabushiki Kaisha Illumination systems and devices for fourier ptychographic imaging
US20170134680A1 (en) * 2014-07-29 2017-05-11 Jinan University Optical imaging method using single pixel detector
CN112923867A (en) * 2021-01-21 2021-06-08 南京工程学院 Fourier single-pixel imaging method based on frequency spectrum significance
CN113284071A (en) * 2021-06-15 2021-08-20 北京理工大学 Periopsis imaging method and system based on human retina mechanism simulation

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170134680A1 (en) * 2014-07-29 2017-05-11 Jinan University Optical imaging method using single pixel detector
WO2016101008A1 (en) * 2014-12-23 2016-06-30 Canon Kabushiki Kaisha Illumination systems and devices for fourier ptychographic imaging
CN112923867A (en) * 2021-01-21 2021-06-08 南京工程学院 Fourier single-pixel imaging method based on frequency spectrum significance
CN113284071A (en) * 2021-06-15 2021-08-20 北京理工大学 Periopsis imaging method and system based on human retina mechanism simulation

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
HUAN CUI等: "Omnidirectional ghost imaging system and unwrapping-free panoramic ghost imaging", OPTICS LETTERS VOL. 46, NO. 22, 15 November 2021 (2021-11-15), pages 5611 - 5614 *

Similar Documents

Publication Publication Date Title
Zhang et al. Fast Fourier single-pixel imaging via binary illumination
CN110057319B (en) A high-speed three-dimensional measurement method for objects with large-scale reflectivity changes
US9961285B2 (en) Optical imaging method using single pixel detector
CN108895985B (en) Object positioning method based on single-pixel detector
CN112184591A (en) An image restoration method based on deep learning image moiré removal
CN108537735B (en) An image stitching method for focal plane terahertz imaging
CN107306333A (en) A kind of high speed single pixel imaging method
CN110333205A (en) Apparatus and method for scattering imaging through a dynamic liquid medium
Liang et al. Fast Fourier single-pixel imaging based on Sierra–Lite dithering algorithm
CN105841925A (en) Detector pixel response Fourier spectrum acquisition-based image reconstruction method
CN108564090B (en) A Fast Fourier Single Pixel Imaging Method Based on Signal Weighting
Huang et al. Compressed Hermite–Gaussian differential single-pixel imaging
CN110942423B (en) A far-field super-resolution reconstruction method based on Fourier stack imaging
López‐Torres et al. Improving 3D reconstruction accuracy in wavelet transform profilometry by reducing shadow effects
CN116132816A (en) Fourier Panoramic Ghost Imaging Method Based on Logarithmic Linear Transform
CN115953334B (en) A Fourier single-pixel imaging reconstruction method based on coordination of time domain and frequency domain
CN111445416B (en) Method and device for generating high-dynamic-range panoramic image
CN113192154A (en) Underwater ghost imaging system based on edge calculation and deep learning image reconstruction method
CN108254917A (en) An edge-enhanced imaging device and method based on fractional photon orbital angular momentum
CN111260747B (en) Method and system for high-throughput optical tomography based on virtual digital modulation
CN103067664B (en) Image compressing and uncompressing method based on optical principle
CN111880389A (en) A method for eliminating zero-order diffraction of infrared digital holography
CN110118755A (en) Optical imaging method based on Fourier's light field and relevance imaging
CN109919892A (en) Image fusion method of filter preset deep learning neural network
Kim et al. DePhaseNet: A deep convolutional network using phase differentiated layers and frequency based custom loss for RGBW image sensor demosaicing

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination