An interference hyperspectral stripe removal method and device

Abstract

The embodiment of the present application discloses an interference hyperspectral stripe removal method and device. The method comprises the following steps: according to the data characteristics of an interference spectrometer, small bilateral interference data of each scene spectral image is extracted; according to the small bilateral interference data, phase information of each frame of bilateral sampling interference data is calculated; the phase frequency information of the phase information of each waveband in space is calculated; the Fourier transform of each waveband of the spectral image in the column direction is carried out, and the high-frequency information position of each frequency spectrum is determined through the frequency of the phase information in space; and the spectral image is processed based on the phase frequency information and the high-frequency information position, so that the spectral image after the stripes are removed is obtained. The embodiment of the present application combines the frequency of the phase and the frequency of the spectral image to suppress the high-frequency information, so that the purpose of removing the image parasitic ghost image is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of aerospace remote sensing, and particularly relates to an interference hyperspectral stripe removal method and device. BACKGROUND

[0002] The hyperspectral imager carried on the environment disaster mitigation satellite No. 2 is a time-space joint modulation interference imaging spectrometer, which generates a three-dimensional image with spatial information and interference information through push-broom on the entire field of view. The spectral distribution of the light source target can be obtained by Fourier transform on the interference diagram, and the hyperspectral image can be synthesized after data image system processing. The time-space joint modulation interference spectrometer has the advantages of potential high throughput, multi-channel and high spectral resolution, and has good development potential and application prospect.

[0003] During the imaging process of the interference spectrometer carried on the environment disaster mitigation satellite No. 2, the CCD (Charge Coupled Device, semiconductor photosensitive element) device will reflect the incident light of the left (right) field of view, thereby generating new interference stripes in the right (left) field of view, and the restored image will be filled with the entire field of view. The parasitic ghost image is caused by the insufficient light antireflection treatment process of the medium interface formed by the silicon material surface of the detector and the air. The light reflected by the detector enters the interferometer again through the Fourier lens, and then reenters the detector after being split by the interferometer and passing through the Fourier lens again, thereby forming new interference data superimposed on the true interference data. The existence of the parasitic ghost image affects the image signal-to-noise ratio and spectral accuracy of the hyperspectral data. SUMMARY

[0004] The present application solves the technical problem of overcoming the deficiencies of the prior art and providing an interference hyperspectral stripe removal method and device.

[0005] The technical solution of the present application is as follows:

[0006] In a first aspect, the present application provides an interference hyperspectral stripe removal method, which comprises the following steps:

[0007] According to the data characteristics of the interference spectrometer, small double-side interference data of each spectral image is extracted;

[0008] According to the small double-side interference data, the phase information of each frame of double-side sampling interference data is calculated;

[0009] The phase frequency information of the phase information of each beam in space is calculated;

[0010] The Fourier transform in the column direction of each waveband of the spectral image is performed, and the high frequency information position of each frequency spectrum is determined through the frequency of the phase information in space;

[0011] processing the spectral image based on the phase frequency information and the high frequency information position, to obtain a spectral image after removing the stripes.

[0012] Optionally, the small bilateral interference data of each scene spectral image is extracted according to characteristics of the interference spectrometer data, and the small bilateral interference data of each scene spectral image comprises:

[0013] The interference data of each scene spectral image is extracted according to characteristics of the interference spectrometer data.

[0014] The initial small bilateral interference data of the interference data is extracted.

[0015] The initial small bilateral interference data is processed by zero padding to obtain the small bilateral interference data of each scene spectral image.

[0016] Optionally, the phase information of each frame bilateral sampling interference data is calculated based on the small bilateral interference data, and the phase information of each frame bilateral sampling interference data comprises:

[0017] The small bilateral interference data is subjected to Fourier transform to obtain complex spectral information.

[0018] The phase information corresponding to each wave number is calculated based on the complex spectral information.

[0019] The phase information of each frame bilateral sampling interference data is calculated based on the phase information corresponding to each wave number.

[0020] Optionally, the spectral image after removing the stripes is obtained by processing the spectral image based on the phase frequency information and the high frequency information position, and the processing comprises:

[0021] The high frequency information of the spectral image is suppressed based on the high frequency information position and the phase frequency information, to obtain the spectral image after removing the stripes.

[0022] In a second aspect, an embodiment of the present application provides an interference hyperspectral stripe removal device, and the device comprises:

[0023] A small bilateral interference data extraction module is configured to extract small bilateral interference data of each scene spectral image according to characteristics of interference spectrometer data.

[0024] A phase information calculation module is configured to calculate phase information of each frame bilateral sampling interference data based on the small bilateral interference data.

[0025] A phase frequency information calculation module is configured to calculate phase frequency information of the phase information in space for each beam.

[0026] a high-frequency information position determination module configured to perform Fourier transform on each waveband of the spectral image in a column direction, and determine a high-frequency information position of each frequency spectrum through the phase information in space;

[0027] a spectral image acquisition module configured to process the spectral image based on the phase frequency information and the high-frequency information position, and obtain a spectral image after removing the fringes.

[0028] Optionally, the small-bilateral interference data extraction module comprises:

[0029] an interference data extraction unit configured to extract interference data of each spectral image according to characteristics of the interference spectrometer data;

[0030] a small-bilateral interference data extraction unit configured to extract initial small-bilateral interference data of the interference data;

[0031] a small-bilateral interference data acquisition unit configured to perform zero padding on the initial small-bilateral interference data, and obtain small-bilateral interference data of each spectral image.

[0032] Optionally, the phase information calculation module comprises:

[0033] a complex spectral information acquisition unit configured to perform Fourier transform on the small-bilateral interference data, and obtain complex spectral information;

[0034] a first phase information calculation unit configured to calculate phase information corresponding to each wave number according to the complex spectral information;

[0035] a second phase information calculation unit configured to calculate phase information of each frame of bilateral sampling interference data based on the phase information corresponding to each wave number.

[0036] Optionally, the spectral image acquisition module comprises:

[0037] a spectral image acquisition unit configured to suppress high-frequency information of the spectral image according to the high-frequency information position and the phase frequency information, and obtain a spectral image after removing the fringes.

[0038] The advantages of this invention compared to existing technologies are as follows: The method and apparatus for removing hyperspectral interference fringes provided in this invention's embodiments are as follows: By extracting small bilateral interference data for each spectral image based on the characteristics of the interferometric spectrometer data, calculating the phase information of each frame of bilateral sampled interference data based on the small bilateral interference data, calculating the spatial phase frequency information of the phase information of each beam, performing a column-direction Fourier transform on each band of the spectral image, and determining the high-frequency information position of each spectrum through the spatial frequency of the phase information, processing the spectral image based on the phase frequency information and the high-frequency information position to obtain the spectral image after removing fringes. This invention's embodiments achieve the purpose of removing parasitic ghost images by combining the frequency of the phase and the frequency of the spectral image to suppress high-frequency information. Attached Figure Description

[0039] Figure 1 A flowchart illustrating the steps of an interference hyperspectral fringe removal method provided in an embodiment of the present invention;

[0040] Figure 2 A schematic diagram of an interference hyperspectral stripe removal process provided in an embodiment of the present invention;

[0041] Figure 3 This is a schematic diagram of the structure of an interference hyperspectral stripe removal device provided in an embodiment of the present invention. Detailed Implementation

[0042] The interferometric data acquired by the Environment-2 interferometric imaging spectrometer contains information from parasitic ghost images, resulting in regular fringe noise in the reconstructed spectral image. To improve the spectral accuracy and image signal-to-noise ratio of the reconstructed image from the interferometric hyperspectral image, based on the actual characteristics of the spectrometer and its imaging principle, high-frequency signal correction was performed on the reconstructed image by combining frequency and phase analysis, thus resolving the parasitic ghost image problem.

[0043] The technical solutions of the embodiments of the present invention will be described in detail below with reference to specific examples.

[0044] Example 1

[0045] Reference Figure 1 The flowchart illustrates the steps of an interference hyperspectral fringe removal method provided by an embodiment of the present invention, as follows: Figure 1 As shown, the method may include the following steps:

[0046] Step 101: Based on the characteristics of the interferometric spectrometer data, extract the small bilateral interference data for each spectral image.

[0047] In the embodiment of the present application, the technical solutions of the embodiment can be described in detail taking the HJ-2 satellite hyperspectral imager as an example.

[0048] In a specific implementation, the small bilateral interference data of each scene spectral image can be extracted according to the data characteristics of the interference spectrometer. Specifically, the interference data of each scene spectral image can be extracted according to the data characteristics of the interference spectrometer; initial small bilateral interference data of the interference data is extracted; the initial small bilateral interference data is processed by zero padding to obtain the small bilateral interference data of each scene spectral image.

[0049] In actual application, the interference data of each scene image (from the first frame to the nth frame, a total of n frames of interference data) can be extracted according to the data characteristics of the interference spectrometer, and the small bilateral interference data (i.e., the initial small bilateral interference data in this example) symmetrically sampled near the zero-order fringe of each frame of data is further extracted.

[0050] Zero padding processing of small bilateral interference data: the zero padding processing of small bilateral interference data is because the sampling points of small bilateral data are few, and the spectral curve wavelength interval after direct spectral restoration is large, which needs to be interpolated, and the zero padding in time domain is equivalent to the interpolation in frequency domain (spectrum and phase information). Thus, the small bilateral interference data of each scene spectral image can be obtained.

[0051] After the small bilateral interference data of each scene spectral image is extracted according to the data characteristics of the interference spectrometer, step 102 is executed.

[0052] Step 102: phase information of each frame of bilateral sampling interference data is calculated according to the small bilateral interference data.

[0053] After the small bilateral interference data of each scene spectral image is extracted according to the data characteristics of the interference spectrometer, the phase information of each frame of bilateral sampling interference data can be calculated according to the small bilateral interference data. Specifically, the small bilateral interference data can be Fourier transformed to obtain complex spectral information; the phase information corresponding to each wavelength is calculated according to the complex spectral information; and the phase information of each frame of bilateral sampling interference data is calculated based on the phase information corresponding to each wavelength.

[0054] In actual application, the detailed steps of calculating the phase information of each frame of bilateral sampling interference data are as follows:

[0055] 1. Calculate the phase error

[0056] The interference data and the spectral data satisfy the Fourier transform relationship, and the bilateral data is Fourier transformed according to (formula 1) to obtain complex spectral information B(v), B(v) is a complex number, B r (v) is a real part, and B i(v) is imaginary part. The phase information corresponding to each wave number is calculated according to (Formula 2);

[0057] B(v) = FT(I(l)) (1)

[0058] Phase(v) = arctan(B i (v) / B r (v)) (2)

[0059] 2. Calculate the phase error of each frame

[0060] Thus, the phase information corresponding to the multi-frame interference data of each scene image can be calculated, wherein the phase information of a column is phase1(v), phase2(v), …, phase n (v).

[0061] After the phase information of each frame of double-side sampling interference data is calculated according to the small double-side interference data, step 103 is performed.

[0062] Step 103: Calculate the phase frequency information of the phase information of each beam in space.

[0063] After the phase information of each frame of double-side sampling interference data is calculated according to the small double-side interference data, the phase frequency information of the phase information of each beam in space (column direction) can be calculated.

[0064] After the phase frequency information of the phase information of each beam in space is calculated, step 104 is performed.

[0065] Step 104: Perform Fourier transform on each waveband of the spectral image in the column direction, and determine the high-frequency information position of each spectrum through the frequency of the phase information in space.

[0066] Fourier transform can express a certain function satisfying certain conditions as a linear combination of trigonometric functions (sine and / or cosine functions) or their integrals. In different research fields, Fourier transform has various different variants, such as continuous Fourier transform and discrete Fourier transform.

[0067] After the phase frequency information of the phase information of each beam in space is calculated, Fourier transform can be performed on each waveband of the spectral image in the column direction, and the high-frequency information position of each spectrum can be determined through the frequency of the phase information in space.

[0068] In a specific implementation, the implementation process is Fourier transform of the restored image, and the high frequency position is determined: Fourier transform is performed on each waveband of the spectral image in the column direction, and the high frequency information position of each spectrum is determined by the frequency of the phase information in the space (column direction).

[0069] After Fourier transform is performed on each waveband of the spectral image in the column direction, and the high frequency information position of each spectrum is determined by the frequency of the phase information in the space, step 105 is performed.

[0070] Step 105: processing the spectral image based on the phase frequency information and the high frequency information position to obtain a spectral image after the stripes are removed.

[0071] After the phase frequency information and the high frequency information position are obtained, the spectral image can be processed based on the phase frequency information and the high frequency information position to obtain a spectral image after the stripes are removed. Specifically, the high frequency information of the spectral image can be suppressed according to the high frequency information position and the phase frequency information to obtain a spectral image after the stripes are removed.

[0072] In actual application, the high frequency information of the spectral image can be suppressed, and image restoration can be performed, that is, the high frequency information of each column is multiplied by 1 / 3, and then each spectrum is taken modulo to restore the spectral image. At this time, a spectral image after the stripes are removed can be obtained, the phase frequency information of the original interference data is fully utilized, and the high frequency information after Fourier transform is removed, so that the purpose of removing image stripe noise is achieved, and the signal-to-noise ratio and spectral accuracy of the final restored spectral data are obviously improved.

[0073] In the spectral restoration process of the interference imaging spectrometer, the stripes of the parasitic ghost image are superimposed on the interference image, so that there are ripples in the entire optical field after the restoration. Through the spatial frequency combination analysis of the phase before the restoration and the image after the restoration, it is concluded that the frequency of the ripple is proportional to the wave number, and the frequency of the ripple is higher and higher with the increase of the wave number. The frequency of the phase in each waveband is consistent with the spatial frequency of the ripple. Therefore, the position of the high frequency information of each waveband of the restored image after Fourier transform can be set by calculating the frequency change of the phase of each spectrum in the spatial scanning direction. The suppression of the high frequency information can achieve the purpose of removing the ripple.

[0074] Next, the technical solutions of the embodiments of the present application are described as follows. Figure 2 The technical solutions of the embodiments of the present application are described as follows.

[0075] Referring to Figure 2 FIG. 1 shows a flowchart of the interference hyperspectral stripe removal process provided by the embodiments of the present application.

[0076] As Figure 2As shown, the interference hyperspectral fringe removal procedure can include the following steps:

[0077] 1. Obtain interference data of each frame of image, such as Figure 2 As shown, the first frame of interference data, the second frame of interference data, the... frame of interference data and the n frame of interference data are obtained, wherein n is a positive integer.

[0078] 2. According to the interference data of each frame, the corresponding phase error of each frame is calculated.

[0079] 3. The phase frequency is calculated according to the phase error of multiple frames.

[0080] 4. The spatial domain FFT processing is performed on the collected spectral data to obtain the high frequency information position.

[0081] 5. The high frequency information in the image is suppressed by combining the phase frequency and the high frequency information position, so that the fringe-free spectral data can be obtained.

[0082] Example two

[0083] Referring to Figure 3 , a structure schematic diagram of an interference hyperspectral fringe removal device provided by the embodiment of the application is shown, as shown in Figure 3 The device can include the following modules:

[0084] The small double-side interference data extraction module 310 is used to extract small double-side interference data of each scene spectral image according to the data characteristics of the interference spectrometer.

[0085] The phase information calculation module 320 is used to calculate the phase information of each frame of double-side sampling interference data according to the small double-side interference data.

[0086] The phase frequency information calculation module 330 is used to calculate the phase frequency information of the phase information in space of each beam.

[0087] The high frequency information position determination module 340 is used to perform Fourier transform on each waveband of the spectral image in the column direction, and determine the high frequency information position of each frequency spectrum through the frequency of the phase information in space.

[0088] The spectral image acquisition module 350 is used to process the spectral image based on the phase frequency information and the high frequency information position, and obtain the spectral image after removing the fringe.

[0089] Optionally, the small double-side interference data extraction module includes:

[0090] The interference data extraction unit is used to extract interference data of each scene spectral image according to the data characteristics of the interference spectrometer.

[0091] a small bilateral interference data extraction unit configured to extract initial small bilateral interference data of the interference data;

[0092] a small bilateral interference data acquisition unit configured to perform zero padding on the initial small bilateral interference data to obtain small bilateral interference data of each spectral image.

[0093] Optionally, the phase information calculation module comprises:

[0094] a complex spectral information acquisition unit configured to perform Fourier transform on the small bilateral interference data to obtain complex spectral information;

[0095] a first phase information calculation unit configured to calculate phase information corresponding to each wave number according to the complex spectral information;

[0096] a second phase information calculation unit configured to calculate phase information of each frame of bilateral sampling interference data based on the phase information corresponding to each wave number.

[0097] Optionally, the spectral image acquisition module comprises:

[0098] a spectral image acquisition unit configured to suppress high-frequency information of the spectral image according to the high-frequency information position and the phase frequency information to obtain a spectral image after removal of the stripes.

[0099] The specific embodiments described in the present application can enable those skilled in the art to have a more comprehensive understanding of the present application, but in no way limit the present application. Therefore, those skilled in the art should understand that modifications or equivalent replacements to the present application can still be made; and all technical solutions and improvements that do not depart from the spirit and technical essence of the present application should be covered in the protection scope of the present application patent.

[0100] The contents not described in detail in the specification of the present application are the known technology of those skilled in the art.

Claims

1. An interferometric hyperspectral striping removal method, characterized in that, The method comprises: According to the characteristics of the interferometric spectrometer data, the interference data of each scene spectral image is extracted; the initial small bilateral interference data of symmetric sampling and less sampling points near the zero-order fringes of each frame of data in the interference data is extracted; the initial small bilateral interference data is zero-filled to obtain the small bilateral interference data of each scene spectral image; According to the small bilateral interference data, the phase information of each frame of bilateral sampling interference data is calculated; The phase frequency information of the phase information of each wave number in space is calculated; The Fourier transform of each waveband of the spectral image in the column direction is performed, and the high-frequency information position of each frequency spectrum is determined through the frequency of the phase information in space; Based on the phase frequency information and the high-frequency information position, the spectral image is processed to obtain a spectral image after removing the fringes.

2. The method of claim 1, wherein, According to the small bilateral interference data, the phase information of each frame of bilateral sampling interference data is calculated, comprising: The Fourier transform of the small bilateral interference data is performed to obtain complex spectral information; According to the complex spectral information, the phase information corresponding to each wave number is calculated; Based on the phase information corresponding to each wave number, the phase information of each frame of bilateral sampling interference data is calculated.

3. The method of claim 1, wherein, Based on the phase frequency information and the high-frequency information position, the spectral image is processed to obtain a spectral image after removing the fringes, comprising: According to the high-frequency information position and the phase frequency information, the high-frequency information of the spectral image is suppressed to obtain a spectral image after removing the fringes.

4. An apparatus for interferometric hyperspectral stripe removal, characterized in that, The device comprises: A small bilateral interference data extraction module is configured to extract the interference data of each scene spectral image according to the characteristics of the interferometric spectrometer data; extract the initial small bilateral interference data of symmetric sampling and less sampling points near the zero-order fringes of each frame of data in the interference data; and perform zero filling on the initial small bilateral interference data to obtain the small bilateral interference data of each scene spectral image; A phase information calculation module is configured to calculate the phase information of each frame of bilateral sampling interference data according to the small bilateral interference data; A phase frequency information calculation module is configured to calculate the phase frequency information of the phase information of each wave number in space; A high-frequency information position determination module is configured to perform the Fourier transform of each waveband of the spectral image in the column direction, and determine the high-frequency information position of each frequency spectrum through the frequency of the phase information in space; A spectral image acquisition module is configured to process the spectral image based on the phase frequency information and the high-frequency information position to obtain a spectral image after removing the fringes.

5. The apparatus of claim 4, wherein, The small bilateral interference data extraction module comprises: An interference data extraction unit is configured to extract the interference data of each scene spectral image according to the characteristics of the interferometric spectrometer data; A small bilateral interference data extraction unit is configured to extract the initial small bilateral interference data of symmetric sampling and less sampling points near the zero-order fringes of each frame of data in the interference data; A small bilateral interference data acquisition unit is configured to perform zero filling on the initial small bilateral interference data to obtain the small bilateral interference data of each scene spectral image.

6. The apparatus of claim 4, wherein, The phase information calculation module comprises: The complex spectral information acquisition unit is configured to perform Fourier transform on the small-bilateral interference data to obtain complex spectral information. The first phase information calculation unit is configured to calculate phase information corresponding to each wave number according to the complex spectral information. The second phase information calculation unit is configured to calculate phase information of each frame of bilateral sampling interference data based on the phase information corresponding to each wave number.

7. The apparatus of claim 4, wherein, The spectral image acquisition module comprises: The spectral image acquisition unit is configured to suppress high-frequency information of the spectral image according to the high-frequency information position and the phase frequency information to obtain a spectral image after removing the stripes.

Images