CN104580937B - A kind of infrared imaging system fringes noise minimizing technology - Google Patents
A kind of infrared imaging system fringes noise minimizing technology Download PDFInfo
- Publication number
- CN104580937B CN104580937B CN201510029313.4A CN201510029313A CN104580937B CN 104580937 B CN104580937 B CN 104580937B CN 201510029313 A CN201510029313 A CN 201510029313A CN 104580937 B CN104580937 B CN 104580937B
- Authority
- CN
- China
- Prior art keywords
- numbers
- ordered series
- image
- noise
- spectral line
- 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.)
- Active
Links
Abstract
The invention discloses a kind of infrared imaging system fringes noise minimizing technology, implementation step is as follows：1) black matrix is imaged using infrared imaging system and the infrared image that will obtain is converted to one-dimensional ordered series of numbers；2) the one-dimensional ordered series of numbers that will be obtained carries out FFT Fourier transformations；3) spectral line that fringes noise is represented in spectrogram after conversion is demarcated；4) result after noise-reduced image FFT is changed for being equally converted to one-dimensional ordered series of numbers is processed with the position of spectral line demarcated；5) result after treatment is inversely transformed into one-dimensional ordered series of numbers, then splices back two dimensional image.It is not enough instant invention overcomes traditional infrared image noise elimination method, there is provided a kind of more effective, more accurate removal fringes noise method.The present invention can effectively remove the fringes noise in infrared image, not lose image prime information, obtain high-quality restored image.
Description
Technical field：
The invention belongs to high sensitivity infrared imaging and applied technical field, based on a kind of refrigeration mode infrared imaging hardware system
System, is related to the removal of the infrared image fringes noise of Fourier transformation.
Background technology：
With the development of infrared imagery technique, requirement more and more higher of the people to infrared imaging system level of sensitivity,
The demand for obtaining high sensitivity high-resolution Thermo-imaging system is also more urgent, by simple Nonuniformity Correction infrared image
, it is necessary to subsequently carry out denoising to infrared image the need for high sensitivity occasion can not be met, to improve the spirit of infrared image
Sensitivity level.High sensitivity infrared imaging system typically uses refrigeration mode infrared focal plane device, and it by matrix form by arranging
Pixel and peripheral reading circuit composition, due to the detection that the first response characteristic of each detection in focal plane arrays (FPA) is not quite identical, different
Light-sensitive element can produce different output signals to identical infra-red radiation so that a kind of striated occurs in infrared image
Heterogeneity noise, conventional Non-uniformity Correction Algorithm cannot remove the noise.In practice frequently with Histogram Matching with
And the method such as match by moment removes such fringes noise, the method for the treatment of is directly to remove noise on 2d.For these
Traditional infrared image noise reduction process method, and be not concerned with noise generation mechanism, but it is simple noise is considered, meeting
Cause the loss of image high-frequency information.It is in fact a kind of one-dimensional signal that the output of focal plane is seen in a sense, equivalent to pressing
Different pixels are in chronological sequence read according to different output ways, so can be produced when these pixels are arranged as into two dimensional image
The fringes noise of light and shade.The frequency of this noise is closely related with the number of the reading circuit of focal plane.This method is based on this problem
And propose, the ardware feature based on infrared imaging, fringes noise is targetedly removed, the premise of image information can not lost
Lower effectively removal fringes noise.
The content of the invention：
Goal of the invention：Instant invention overcomes the deficiency of conventional method, there is provided a kind of more effective, more accurate removal striped is made an uproar
The method of sound.Infrared two dimensional image is converted to one-dimensional discrete signal and carries out FFT treatment by this method, from the ardware feature of system
The method that proposition of setting out removes noise.
Infrared imaging system fringes noise minimizing technology proposed by the present invention, it is characterised in that comprise the following steps：
1) using the infrared imaging system of the focal plane arrays (FPA) based on M × N scales, uniform black matrix is imaged,
The image picture elements that will be obtained according to putting in order that P roads reading circuit is exported, with 1,2,3 ..., in chronological sequence arrange by the order on P roads
Arrange into one-dimensional, obtain I (N) ordered series of numbers.Wherein M is the row number of focal plane, and wherein M is the integral multiple of P.
2) FFT is carried out to the one-dimensional discrete data obtained in 1)：
F=FFT (I (N))
W=abs (F)
With sample frequency as A, (A should be greater than N, M, and the precision in occurrence influence step 4, representative value takes length (I)) does
Go out spectrogram.
Wherein FFT () represents quick one-dimensional Fourier transform, and abs () represents the extraction function of ordered series of numbers I frequency domain amplitudes.
3) to the image of the frequency of W ordered series of numbers that obtains in 2) and amplitude, X-axis with following formula as coordinate (1~length (I)) ×
A/length (I), Y-axis coordinate amplitude is W.Now obtain frequency present in the one-dimensional ordered series of numbers that Infra-Red Imaging System is imaged to black matrix
Rate spectrum.Wherein length (I) is the length of ordered series of numbers I.
4) to the spectrum obtained in 3), it is the cycle that its spectral line Y-axis amplitude correspond in ordered series of numbers I with (A/x) individual pixel
Fringes noise intensity.Wherein x is the corresponding X-axis coordinate of spectral line.
5) it is very high to amplitude in the spectrum in 3) and corresponding in ordered series of numbers I with P, M/n (n is 1,2,3 ...) and M pixel
It is the spectral line in cycle, records its X-axis coordinate x1, x2, x3 ..., x1*, x2*, x3* ..., wherein x1, x1* P pixel of correspondence is week
The spectral line of phase, by that analogy.Then the position of these spectral lines correspond to F ordered series of numbers in position of spectral line for (x1, x2, x3 ... x1*,
x2*,x3*…)/A×length(I)。
6) infrared image of noise reduction is treated in input, as shown in 1), is changed into one-dimensional ordered series of numbers and is carried out FFT conversions, in 5)
The position of spectral line for obtaining is by corresponding value zero setting in FFT transformation results.
7) by the ordered series of numbers FFT inverse transformations after zero setting in 6), and one-dimensional ordered series of numbers is spelled into back original X-Y scheme according to the rule in 1)
Picture, obtains removing the infrared image after fringes noise.
The present invention has the following advantages and effect relative to existing method：
Hardware point of view of the present invention from system, it is contemplated that fringes noise Producing reason, and targetedly by two dimension
Image is changed into one dimensional image carries out noise reduction, can more effectively extract noise, burbling noise and useful signal, it is to avoid to two
Blindness during dimension image procossing.
The present invention is demarcated with black matrix image to the spectrum position of fringes noise, and can be commonly used to follow-up image drop
Make an uproar, improve the accuracy of noise reduction.
The present invention goes the mode of spectral line to replace traditional filtering mode using Fourier.Can targetedly remove by reading
The fringes noise that circuit is produced, and the HFS of image has been fully retained, the method can more effectively remove noise, and completely
Retain image detail.
Brief description of the drawings：
Fig. 1 is the general frame that the infrared image based on Fourier transformation of the invention removes fringes noise method.
Fig. 2 is the contrast of black matrix image (b) after original black matrix image (a) and treatment in the embodiment of the present invention.
Fig. 3 be black matrix image in the embodiment of the present invention spectrum (a) artwork with treatment after (b) contrast.
Fig. 4 is the contrast that noise-reduced image (a) and image (b) after treatment are treated in the embodiment of the present invention.
Fig. 5 is the treatment image finally given in the embodiment of the present invention.
Specific embodiment：
With reference to embodiment and accompanying drawing, the present invention is described in further detail, but embodiments of the present invention are not limited
In this.
Such as Fig. 1, so that 320 × 256 focal planes based on No. 4 reading circuits outputs are for the infrared imaging system of hardware as an example,
Improved infrared system fringes noise minimizing technology, comprises the following steps：
1) uniform black matrix is imaged using system.The image picture elements that will be obtained are suitable according to the arrangement that No. four reading circuits are exported
Sequence, with 1, the order on 2,3,4 tunnels is in chronological sequence arranged in one-dimensional, obtains I (N) ordered series of numbers.320 is columns.
2) FFT is carried out with sample frequency as 81920 to the one-dimensional discrete data obtained in 1)：
F=FFT (I (N))
W=abs (F)
Wherein FFT () represents quick one-dimensional Fourier transform, and abs () represents the extraction function of ordered series of numbers I frequency domain amplitudes.
3) image of frequency and amplitude is made to the W ordered series of numbers obtained in 2), X-axis coordinate is 1:length(I))×81920/
Length (I)), Y-axis coordinate is W.Now obtain frequency spectrum present in the one-dimensional ordered series of numbers that Infra-Red Imaging System is imaged to black matrix
Line chart.Wherein length (I) is the length of ordered series of numbers I.
4) to the spectrum obtained in 3), it is the striped in cycle that its spectral line correspond in ordered series of numbers I with (81920/x) individual pixel
Noise.Wherein x is the corresponding X-axis coordinate of spectral line.
5) it is very high to amplitude in the spectrum in 3) and corresponding to 4,160,320 pixels being the spectrum in cycle in ordered series of numbers I
Line, records its X-axis coordinate 257,509,20481,81665,81413,61441.
6) infrared image of noise reduction is treated in input, as shown in 1), is changed into one-dimensional ordered series of numbers and is carried out FFT conversions, in 5)
The position of spectral line for obtaining is by corresponding value zero setting in FFT transformation results.
7) by the ordered series of numbers FFT inverse transformations after zero setting in 6), and one-dimensional ordered series of numbers is spelled into back former two dimensional image, obtains removing striped
Infrared image after noise.
It is that the black matrix image local not processed by context of methods amplifies, wherein can be obvious as shown in Fig. 2 (a)
See and there is striated noise, Fig. 2 (b) is that, by the image after context of methods treatment, fringes noise is eliminated substantially.Figure
3 is the change contrast of spectrum before and after black matrix image procossing, it can be seen that fringes noise is mainly too high by several amplitudes
Frequency corresponding to spectral line is caused.
With this position of spectral line as standard, noise reduction is carried out to outdoor scene image, what such as Fig. 4 (b) showed the present embodiment removes striped
As a result, Fig. 4 (b) is compared with the grandfather tape stripe pattern in Fig. 4 (a), it is seen then that striped therein is effectively removed, rich
Rich image detail, at the same the details of image especially marginal portion has obtained complete reservation, picture quality greatly improved.
Fig. 5 is the image adjusted by gray scale after processing.
The preferred embodiment of the present invention is the foregoing is only, protection scope of the present invention is not limited in above-mentioned embodiment party
Formula, every technical scheme for belonging to the principle of the invention belongs to protection scope of the present invention.For those skilled in the art
Speech, some improvements and modifications carried out on the premise of principle of the invention is not departed from, these improvement also should be regarded as of the invention
Protection domain.
Claims (1)
1. a kind of infrared imaging system fringes noise minimizing technology, it is characterised in that comprise the following steps：
1) using the infrared imaging system of the focal plane arrays (FPA) based on M × N scales, uniform black matrix is imaged, will be obtained
Image picture elements according to putting in order that P roads reading circuit is exported, P is the way of ROIC for FPA output, with 1,
The order on 2,3 ... P roads is in chronological sequence arranged in one-dimensional, obtains I (N) ordered series of numbers, and wherein M is the row number of focal plane, and wherein M is
The integral multiple of P；
2) to step 1) in the one-dimensional discrete data that obtain carry out FFT：
F=FFT [I (N)] (1)
W=abs (F) (2)
With sample frequency as A, A should be greater than N, M, and representative value takes length (I) and makes spectrogram；Wherein FFT () represents quick one
Dimension Fourier transformation, abs () represents the extraction function of ordered series of numbers I frequency domain amplitudes；
3) to step 2) in the frequency of W ordered series of numbers that obtains and amplitude image, X-axis with following formula as coordinate (1~length (I)) ×
A/length (I), Y-axis amplitude coordinate is W, now obtains frequency present in the one-dimensional ordered series of numbers that Infra-Red Imaging System is imaged to black matrix
Rate spectrum, wherein length (I) are the length of ordered series of numbers I (N)；
4) to step 3) in the spectrum that obtains, it is the cycle that its spectral line Y-axis amplitude correspond in ordered series of numbers I (N) with (A/x) individual pixel
Fringes noise intensity, wherein x be the corresponding X-axis coordinate of spectral line；
5) to step 3) in spectrum in amplitude it is very high and corresponding to P, M/n, n being 1,2,3 in ordered series of numbers I (N) ..., and M
Pixel is the spectral line in cycle, records its X-axis coordinate x1, x2, x3 ..., and x1*, 2*, x3* ..., wherein x1, x1* P pixel of correspondence are
The spectral line in cycle, by that analogy；Then the position of these spectral lines correspond to F ordered series of numbers in position of spectral line for (x1, x2, x3 ... x1*,
x2*,x3*…)/A×length(I)；
6) infrared image of noise reduction, such as step 1 are treated in input) as shown in, it is changed into one-dimensional ordered series of numbers and carries out FFT conversions, with step
5) position of spectral line obtained in is by corresponding value zero setting in this FFT transformation result；
7) by step 6) in ordered series of numbers FFT inverse transformations after zero setting, and by one-dimensional ordered series of numbers according to step 1) in rule spells back original two
Dimension image, obtains removing the infrared image after fringes noise.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510029313.4A CN104580937B (en) | 2015-01-21 | 2015-01-21 | A kind of infrared imaging system fringes noise minimizing technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510029313.4A CN104580937B (en) | 2015-01-21 | 2015-01-21 | A kind of infrared imaging system fringes noise minimizing technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104580937A CN104580937A (en) | 2015-04-29 |
| CN104580937B true CN104580937B (en) | 2017-06-27 |
Family
ID=53096031
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510029313.4A Active CN104580937B (en) | 2015-01-21 | 2015-01-21 | A kind of infrared imaging system fringes noise minimizing technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104580937B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106462957B (en) * | 2016-05-19 | 2019-03-08 | 深圳大学 | The minimizing technology and system of fringes noise in a kind of infrared image |
| CN108665422B (en) * | 2017-08-30 | 2022-05-10 | 西安电子科技大学 | Single-frame infrared heterogeneity detection method based on reverse sensing in Fourier domain |
| CN108305220B (en) * | 2017-12-29 | 2020-06-02 | 华中科技大学 | Airborne infrared degraded image correction method |
| CN110101362B (en) * | 2019-04-29 | 2021-12-17 | 中国科学院宁波工业技术研究院慈溪生物医学工程研究所 | Method for removing image noise related to OCT and OCTA |
| CN110689494B (en) * | 2019-09-23 | 2020-10-20 | 北京航天宏图信息技术股份有限公司 | Stripe removing method and device, electronic equipment and storage medium |
| CN111311510B (en) * | 2020-01-22 | 2023-08-25 | 京东方科技集团股份有限公司 | Method and device for removing stripe noise in image |
| CN115272137B (en) * | 2022-09-28 | 2022-12-20 | 北京万龙精益科技有限公司 | Real-time fixed pattern noise removing method, device, medium and system based on FPGA |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1997001153A1 (en) * | 1995-06-23 | 1997-01-09 | Philips Electronics N.V. | Image processing for noise reduction |
| JP2000350096A (en) * | 1999-06-03 | 2000-12-15 | Tech Res & Dev Inst Of Japan Def Agency | Sensitivity correction method |
| CN1917590A (en) * | 2005-05-26 | 2007-02-21 | 红外线解决方案公司 | Method for fixed pattern noise reduction in infrared imaging cameras |
| CN103500443A (en) * | 2013-10-10 | 2014-01-08 | 中国科学院上海技术物理研究所 | Fourier transform-based infrared polarization image fusion method |
-
2015
- 2015-01-21 CN CN201510029313.4A patent/CN104580937B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1997001153A1 (en) * | 1995-06-23 | 1997-01-09 | Philips Electronics N.V. | Image processing for noise reduction |
| JP2000350096A (en) * | 1999-06-03 | 2000-12-15 | Tech Res & Dev Inst Of Japan Def Agency | Sensitivity correction method |
| CN1917590A (en) * | 2005-05-26 | 2007-02-21 | 红外线解决方案公司 | Method for fixed pattern noise reduction in infrared imaging cameras |
| CN103500443A (en) * | 2013-10-10 | 2014-01-08 | 中国科学院上海技术物理研究所 | Fourier transform-based infrared polarization image fusion method |
Non-Patent Citations (4)
| Title |
|---|
| THz图像条纹噪声消除方法研究;邹园园;《中国优秀硕士学位论文全文数据库》;20091015(第10期);全文 * |
| 基于傅里叶变换的条带噪声去除方法;杜艺，刘文国，葛帅;《测绘与空间地理信息》;20140831;第37卷(第8期);全文 * |
| 红外图像条带噪声消除算法研究;王阳;《中国优秀硕士学位论文全文数据库》;20121215(第S12期);全文 * |
| 红外图像条纹噪声消除方法;隋修宝;《红外与毫米波学报》;20120430;第31卷(第2期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104580937A (en) | 2015-04-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104580937B (en) | A kind of infrared imaging system fringes noise minimizing technology | |
| CN110023810B (en) | Digital correction of optical system aberrations | |
| CN110310237B (en) | Method and system for removing image moire, measuring brightness of display panel sub-pixel point and repairing Mura defect | |
| CN107071234B (en) | Lens shadow correction method and device | |
| CN105761231B (en) | A method of for removing fringes noise in high-resolution astronomy image | |
| CN110517206B (en) | Method and device for eliminating color moire | |
| CN107727231B (en) | Method for correcting stripe noise of spectral imager | |
| JP2010166513A (en) | Image processor and digital camera | |
| Hu et al. | Source camera identification using large components of sensor pattern noise | |
| CN106296608B (en) | Mapping table-based fisheye image processing method and system | |
| CN106651962A (en) | Adaptive light compensation method based on projector-camera system | |
| CN113763271A (en) | High-quality spectrum denoising method based on physical noise model | |
| CN105869129A (en) | Residual heterogeneous noise elimination method for aiming at thermal infrared image after heterogeneous correction | |
| CN111242884B (en) | Image dead pixel detection and correction method and device, storage medium and camera equipment | |
| CN115294149A (en) | Astronomical image background extraction method and system | |
| JP2015177244A (en) | Image processing apparatus and image processing method | |
| Raimundo et al. | LAN: Lightweight attention-based network for RAW-to-RGB smartphone image processing | |
| CN110852947B (en) | Infrared image super-resolution method based on edge sharpening | |
| CN103679648B (en) | A kind of match by moment satellite image Strip noise removal method based on space segmentation | |
| CN107862670B (en) | Image restoration method for infrared imaging electric crosstalk | |
| CN111652820B (en) | Method, device and image processing system for correcting non-uniformity of infrared image stripes | |
| CN110017898B (en) | High-precision wavelength calibration method based on interference imaging spectrometer | |
| Liu et al. | Noise analysis and image restoration for optical sparse aperture systems | |
| CN112565637B (en) | Method for removing stripe noise under low illumination in monochromatic sCMOS camera | |
| JP6552520B2 (en) | Calibration method for calibrating a digital imaging device and the digital imaging device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |