CN111080561A - Time domain high-pass filtering method - Google Patents

Time domain high-pass filtering method Download PDF

Info

Publication number
CN111080561A
CN111080561A CN201911088125.3A CN201911088125A CN111080561A CN 111080561 A CN111080561 A CN 111080561A CN 201911088125 A CN201911088125 A CN 201911088125A CN 111080561 A CN111080561 A CN 111080561A
Authority
CN
China
Prior art keywords
image
infrared image
filtering
images
pass filtering
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
CN201911088125.3A
Other languages
Chinese (zh)
Inventor
郑循江
叶志龙
颜志强
孙朔冬
高原
何峰
武斌
董时
吴迪
胡海泉
王秉文
武少冲
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.)
Shanghai Aerospace Control Technology Institute
Original Assignee
Shanghai Aerospace Control Technology Institute
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 Shanghai Aerospace Control Technology Institute filed Critical Shanghai Aerospace Control Technology Institute
Priority to CN201911088125.3A priority Critical patent/CN111080561A/en
Publication of CN111080561A publication Critical patent/CN111080561A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/10—Image enhancement or restoration using non-spatial domain filtering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/70—Denoising; Smoothing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10048—Infrared image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20172—Image enhancement details
    • G06T2207/20192—Edge enhancement; Edge preservation

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Image Processing (AREA)

Abstract

A time domain high-pass filtering method includes the steps of conducting preliminary filtering on original infrared images by using a mask plate, conducting difference value on two adjacent frames of infrared images after preliminary filtering to obtain difference value images, conducting bilateral filtering on the infrared images after preliminary filtering to obtain residual images after bilateral filtering, enabling a part, with a gray value larger than a gray threshold, of the difference value images to belong to a fast moving area, enabling a part, with a gray value smaller than or equal to the gray threshold, of the difference value images to belong to a slow moving area, respectively determining correction terms of the fast moving area and the slow moving area according to the residual images after bilateral filtering and different time constants, and correcting the original infrared images by using the correction terms to obtain corrected infrared images. The invention effectively reduces the non-uniformity of the infrared image, inhibits the occurrence of ghost images and has good correction effect.

Description

Time domain high-pass filtering method
Technical Field
The invention relates to the field of infrared image non-uniformity correction methods, in particular to an improved time domain high-pass filtering method based on adjacent frame difference values.
Background
The existing infrared imaging system mostly uses an infrared focal plane array as a core device, and has the advantages of high detection sensitivity, small system volume, compact structure and the like, so that the infrared imaging system is widely applied to the military and civil fields. However, the quality of the infrared image is affected by the non-uniformity of the infrared focal plane array, and the poor non-uniformity not only weakens the edge of the detected target in the infrared image, but also reduces the definition of the image, and even causes image distortion when the image is serious. The non-uniformity of the infrared focal plane array prevents the further improvement of the quality of the infrared image, and limits the application and development of an infrared imaging system to a certain extent, so that the research on the non-uniformity correction method of the infrared focal plane array is urgently needed.
The infrared image non-uniformity correction algorithm is mainly divided into two types: one is a black body calibration based non-uniformity correction algorithm, and the other is a scene-based infrared image non-uniformity correction algorithm. The correction parameter value in the nonuniformity correction algorithm based on black body calibration is fixed, and the nonuniformity of the infrared focal plane array during working is influenced by the working environment, the working temperature and other factors, so if the initial correction parameter is still adopted to carry out nonuniformity correction on the infrared image, the nonuniformity under the infrared image cannot be effectively improved. If the infrared image is ensured to have better non-uniformity, the correction parameters need to be measured again, and the working process is complicated. In contrast, the infrared image nonuniformity correction algorithm based on the scene does not need to consider the calibration of the correction parameters, updates the correction parameters by using the changed scene information, can automatically replace the correction parameters according to the change of the environment where the infrared system is located, gets rid of manual intervention, and solves the drift problem of the nonuniformity of the focal plane of the nonuniformity correction algorithm based on the black body calibration along with the change of time and temperature.
The infrared image non-uniformity correction algorithm based on the scene is various, and mainly comprises a time domain high-pass filtering algorithm, a least mean square correction algorithm, a constant statistical method, a non-uniformity correction algorithm based on image registration and the like. The time domain high-pass filtering algorithm is a typical infrared image non-uniformity correction algorithm, and the algorithm principle is as follows: in the time domain, the non-uniform noise belongs to low-frequency components, the scene belongs to high-frequency components, and the non-uniform correction is realized by filtering the low-frequency components in the image. However, the long-term still object is easily regarded as the non-uniformity information and then filtered out during filtering, and meanwhile, the sudden change of the object motion state also causes the corrected image to generate a 'ghost image', and the algorithm also has a 'blind area' for non-uniformity correction. Typical improved time-domain high-pass filtering methods include a bilateral filtering-based time-domain high-pass filtering algorithm (BFTH algorithm) and a weighted guided filtering-based time-domain high-pass filtering algorithm (WGTH algorithm). The high-frequency region in the infrared image corrected by the BFTH algorithm still has residual non-uniform noise, and then a ghost image is generated. Although the WGTH algorithm can better solve the problem of ghost image existing in the traditional time domain high-pass filtering algorithm, the non-uniform noise cannot be completely removed.
Disclosure of Invention
The invention provides a time domain high-pass filtering method, which effectively reduces the non-uniformity of an infrared image, inhibits the occurrence of ghost images and has good correction effect.
In order to achieve the above object, the present invention provides a time domain high-pass filtering method, which includes performing preliminary filtering on an original infrared image by using a mask, performing a difference value on two adjacent frames of preliminarily filtered infrared images to obtain a difference image, performing bilateral filtering on the preliminarily filtered infrared image to obtain a residual image after bilateral filtering, enabling a part of the difference image with a gray value greater than a gray threshold to belong to a fast moving region, enabling a part of the difference image with a gray value less than or equal to the gray threshold to belong to a slow moving region, determining correction terms of the fast moving region and the slow moving region according to the residual image after bilateral filtering and different time constants, and correcting the original infrared image by using the correction terms to obtain a corrected infrared image.
The temporal high-pass filtering method according to claim 1, wherein said correction term is:
Figure BDA0002266045450000021
wherein n represents the frame order, n is 2, 3, …, MminIs a small time constant, MmaxIs a large time constant and is therefore,
Figure BDA0002266045450000022
is a residual image after bilateral filtering,
Figure BDA0002266045450000023
Figure BDA0002266045450000024
representing the preliminary filtered infrared image,
Figure BDA0002266045450000025
Xnrepresenting the n-th frame of original infrared image, MASK representing MASK, IrnIs a difference image of the object,
Figure BDA0002266045450000026
Irn(i, j) is the pixel value of the ith row and the jth column in the difference image, and Th is the grayscale threshold.
And selecting the average value of each frame of original infrared image as a gray threshold value.
Said small time constant MminA value of 20, the large time constant MmaxThe value is 40.
And correcting the original infrared image by using a correction term:
Yn＝Xn-fn
wherein, YnRepresenting the corrected output image of the nth frame, XnRepresenting the n-th original infrared image, fnIs the correction term.
The invention uses the mask plate to effectively filter the original infrared image, achieves the purpose of enhancing the edge information of the infrared image, then divides the image into a moving area and a static area according to the set gray threshold value, the moving area selects a smaller time constant, the static area selects a larger time constant, and the moving area and the static area are respectively subjected to infrared non-uniformity correction by combining with an improved correction item, thereby effectively reducing the non-uniformity of the infrared image, simultaneously inhibiting the occurrence of ghost images and having good correction effect.
Drawings
Fig. 1 is a flow chart of a time-domain high-pass filtering method provided by the present invention.
FIG. 2 is an image of a real-shot infrared image using the time-domain high-pass filtering method provided by the present invention.
FIG. 3 is an image of a simulated infrared image using the time domain high-pass filtering method provided by the present invention.
Fig. 4 is an image of a real-shot infrared image corrected by a bilateral filtering-based time domain high-pass filtering algorithm (BFTH).
Fig. 5 is an image of a simulated infrared image corrected by a bilateral-filtering-based time domain high-pass filtering algorithm (BFTH).
Fig. 6 is an image of a live-shot infrared image corrected by a time-domain high-pass filtering algorithm (WGTH) based on weighted guided filtering.
Fig. 7 is an image of a simulated infrared image corrected using a weighted guided filtering based time domain high pass filtering algorithm (WGTH).
Fig. 8 is a roughness curve of a live infrared image after the nonuniformity correction.
Fig. 9 is a roughness curve of a simulated infrared image after non-uniformity correction.
Fig. 10 is a peak signal-to-noise ratio (PSNR) curve of a simulated infrared image after non-uniformity correction.
Fig. 11 is a Structural Similarity (SSIM) curve of a simulated infrared image after non-uniformity correction.
Detailed Description
The preferred embodiment of the present invention will be described in detail below with reference to fig. 1 to 11.
Considering that the traditional time domain high-pass filtering algorithm can filter the edge information of the image while filtering the non-uniformity of the infrared image, the mask plate can be used for filtering the original infrared image to achieve the purpose of enhancing the edge information of the image. In addition, the main disadvantages of the conventional time-domain high-pass filtering algorithm are as follows: the algorithm has poor correction effect on static areas and highlight areas in image scenes, and simultaneously the corrected infrared images also have ghost images. The method for solving the problems is as follows: the infrared image is divided into areas with different motion states, and the areas with different motion states adopt different non-uniformity correction modes. The concrete description is as follows: the method comprises the steps of obtaining a difference image by making a difference value between adjacent frame images, introducing a gray threshold value, selecting a proper threshold value, considering a part of the difference image with the gray value larger than the threshold value as a motion area, considering a correction term depending on a pre-correction term and a bilateral filtered residual image, selecting a smaller time constant, considering a part of the difference image with the gray value smaller than the threshold value as a static area, considering a correction term depending on the pre-correction term and the bilateral filtered residual image, and selecting a larger time constant.
As shown in fig. 1, the present invention provides a time domain high-pass filtering method, which is an improved time domain high-pass filtering algorithm based on an adjacent frame difference, and specifically includes the following steps:
step S1, using a mask plate to perform image processing on the original infrared image XnCarrying out preliminary filtering;
Figure BDA0002266045450000041
wherein, XnRepresenting the n-th frame of original infrared image, MASK representing a MASK plate in a matrix form,
Figure BDA0002266045450000042
representing the primarily filtered infrared image, n represents the frame order, and n is 2, 3, …;
step S2, performing difference on the two adjacent infrared images after the preliminary filtering to obtain a difference image Irn；
Figure BDA0002266045450000043
Step S3, selecting proper gray threshold Th, and taking difference image IrnDivided into two parts, a difference image IrnThe part with middle gray value greater than the threshold Th belongs to the fast moving area and the difference image IrnThe part of the middle gray value less than or equal to the gray threshold Th belongs to the slow moving area;
selection principle of the gray threshold Th: the selection of the gray threshold Th determines the accuracy of division of the moving area and the static area in the infrared image, if the value of the gray threshold Th is too large, part of the moving area in the infrared image is also regarded as the static area, and therefore a larger time constant M is selectedmaxEasily causing the appearance of "ghost images"; if the value of the gray threshold Th is too small, a part of static area in the infrared image is regarded as a motion area, so that a smaller time constant M is selectedminFurther, infrared inhomogeneity cannot be completely removed;
s4, carrying out bilateral filtering on the infrared image after the preliminary filtering to obtain an infrared image after bilateral filtering
Figure BDA0002266045450000051
And bilateral filtered residual image
Figure BDA0002266045450000052
Figure BDA0002266045450000053
Step S5, determining correction terms f of the fast movement area and the slow movement area respectivelyn；
For fast motion regions (motion regions) whose correction term depends on the pre-correction term and the bilateral filtered residual image, a smaller time constant M may be chosenmin(ii) a For slow moving regions (stationary regions), the correction term also depends on the pre-correction term and the bilateral filtered residual image, but a larger time constant M is chosenmax(ii) a Time constant MminAnd MmaxThe empirical parameters are selected according to the image sequence condition;
Figure BDA0002266045450000054
and step S6, correcting the original infrared image by using the correction term.
Yn＝Xn-fn；
Wherein, YnIndicating the corrected output image of the nth frame.
The effect of the improved time domain high-pass filtering algorithm based on the adjacent frame difference is verified by selecting two groups of infrared image sequences, wherein the two groups of infrared image sequences are respectively as follows: a sequence of 1000 infrared images taken and a sequence of 500 infrared images simulated. The MASK plate MASK is in a matrix form, and the size of the MASK plate MASK is [ 111; 181; 111]/16. Selecting the XnTaking the mean value of the frame images as the XthnThe gray threshold Th of the frame image. Smaller time constant MminValue 20, larger time constant MmaxThe value is 40.
As shown in fig. 2 and fig. 3, the image edge is more complete after the improved temporal high-pass filtering algorithm based on the adjacent frame difference provided by the present invention is adopted, the contrast of the image is almost unchanged, and simultaneously, no ghost image is generated. As shown in a) of fig. 3, the image texture hierarchy is relatively complete after the improved temporal high-pass filtering algorithm based on the adjacent frame difference provided by the present invention is adopted, and the brightness of the image is relatively well maintained. As shown in fig. 3, diagram d), no "ghost image" appears in the image.
And as a comparison analysis, introducing a bilateral filtering-based time domain high-pass filtering algorithm. As shown in fig. 4 and 5, the contrast of the image corrected by the temporal high-pass filtering algorithm based on bilateral filtering is reduced, and the image is concentrated in the portrait portion of the image. As can be seen in the lower left corner of the graph b) in fig. 4, the image corrected by the temporal high-pass filtering algorithm based on bilateral filtering loses part of the details, but the corrected image does not generate "ghost images". As shown in fig. 5, in the graphs c) and d), for more serious non-uniform noise in the image, a good correction effect can be achieved by using a bilateral filtering-based time-domain high-pass filtering algorithm. As shown in fig. 5, a) and c), for a relatively serious stripe noise existing in the simulated infrared image sequence, the time domain high-pass filtering algorithm based on bilateral filtering has a good improvement effect, and details in the image are clearly visible.
As a contrast analysis, a time domain high-pass filtering algorithm based on weighted guided filtering is introduced. As shown in fig. 6 and 7, the contrast of the image corrected by the time-domain high-pass filtering algorithm based on the weighted guided filtering is basically unchanged, the brightness of the person region in the real-shot infrared image is still high, the details in the image are relatively complete, the non-uniformity of the simulated infrared image is almost completely corrected, the details of the corrected image are visible, and no ghost image appears, but the algorithm is complex and has poor timeliness.
And evaluating the performance of the algorithm by adopting three evaluation indexes of roughness, peak signal-to-noise ratio and structural similarity. The lower the roughness value is, the lower the non-uniformity degree of the image is, the larger the peak signal-to-noise ratio value is, the lower the noise level of the image is, and the closer the structural similarity value is to 1, the more complete the image detail preservation is.
As shown in fig. 8 and 9, compared with the time domain high-pass filtering algorithm based on bilateral filtering and the time domain high-pass filtering algorithm based on weighted guided filtering, the improved time domain high-pass filtering algorithm based on the adjacent frame difference provided by the present invention has a lower roughness value and better stability.
As can be seen from fig. 10, for the simulated infrared image, the peak signal-to-noise ratio of the image after the improved time domain high-pass filtering algorithm based on the adjacent frame difference provided by the present invention is much larger than the peak signal-to-noise ratio of the image after the time domain high-pass filtering algorithm based on the bilateral filtering, which is improved by about 9% on average.
As can be seen from fig. 11, for the simulated infrared image, compared with the time domain high-pass filtering algorithm based on bilateral filtering and the time domain high-pass filtering algorithm based on weighted guided filtering, the improved time domain high-pass filtering algorithm based on the adjacent frame difference provided by the present invention has a relatively high numerical value of the structural similarity of the image after the present invention is used, which is higher by about 5%.
The invention has the beneficial effects that: the method comprises the steps of using a mask plate to effectively filter an original infrared image, achieving the purpose of enhancing edge information of the infrared image, dividing the image into a moving area and a static area according to a set gray threshold, selecting a smaller time constant for the moving area and a larger time constant for the static area, and respectively carrying out infrared non-uniformity correction on the moving area and the static area by combining an improved correction item, so that the non-uniformity of the infrared image is effectively reduced, the occurrence of ghost images is restrained, and the correction effect is good.
While the present invention has been described in detail with reference to the preferred embodiments, it should be understood that the above description should not be taken as limiting the invention. Various modifications and alterations to this invention will become apparent to those skilled in the art upon reading the foregoing description. Accordingly, the scope of the invention should be determined from the following claims.

Claims (5)

1. A time domain high-pass filtering method is characterized in that a mask plate is used for conducting preliminary filtering on an original infrared image, difference values are conducted on two adjacent frames of infrared images after the preliminary filtering to obtain a difference image, bilateral filtering is conducted on the infrared image after the preliminary filtering to obtain a residual image after the bilateral filtering, a part of the difference image, with a gray value larger than a gray threshold value, belongs to a fast moving area, a part of the difference image, with a gray value smaller than or equal to the gray threshold value, belongs to a slow moving area, correction terms of the fast moving area and the slow moving area are determined according to the residual image after the bilateral filtering and different time constants respectively, and the original infrared image is corrected by the correction terms to obtain a corrected infrared image.
2. The temporal high-pass filtering method according to claim 1, wherein said correction term is:
Figure FDA0002266045440000011
wherein n represents the frame order, n is 2, 3, …, MminIs a small time constant, MmaxIs a large time constant and is therefore,
Figure FDA0002266045440000012
is a residual image after bilateral filtering,
Figure FDA0002266045440000013
Figure FDA0002266045440000014
representing the preliminary filtered infrared image,
Figure FDA0002266045440000015
Xnrepresenting the n-th frame of original infrared image, MASK representing MASK, IrnIs a difference image of the object,
Figure FDA0002266045440000016
Irn(i, j) is the pixel value of the ith row and the jth column in the difference image, and Th is the grayscale threshold.
3. The temporal high-pass filtering method according to claim 2, wherein the mean value of the original infrared image of each frame is selected as the threshold value of the gray scale.
4. The temporal high-pass filtering method according to claim 2, characterized in that said small time constant MminA value of 20, the large time constant MmaxThe value is 40.
5. The temporal high-pass filtering method according to claim 2, characterized in that the correction term is used to correct the original infrared image:
Yn＝Xn-fn
wherein, YnRepresenting the corrected output image of the nth frame, XnRepresenting the n-th original infrared image, fnIs the correction term.
CN201911088125.3A 2019-11-08 2019-11-08 Time domain high-pass filtering method Pending CN111080561A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911088125.3A CN111080561A (en) 2019-11-08 2019-11-08 Time domain high-pass filtering method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911088125.3A CN111080561A (en) 2019-11-08 2019-11-08 Time domain high-pass filtering method

Publications (1)

Publication Number Publication Date
CN111080561A true CN111080561A (en) 2020-04-28

Family

ID=70310738

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911088125.3A Pending CN111080561A (en) 2019-11-08 2019-11-08 Time domain high-pass filtering method

Country Status (1)

Country Link
CN (1) CN111080561A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111968066A (en) * 2020-08-27 2020-11-20 烟台艾睿光电科技有限公司 Infrared image correction method, device and equipment and refrigeration infrared imaging system
CN112651886A (en) * 2020-11-25 2021-04-13 杭州微帧信息科技有限公司 Method for removing color bands in enhanced image by combining original image
CN113349810A (en) * 2021-05-27 2021-09-07 北京安德医智科技有限公司 Cerebral hemorrhage focus identification and hematoma expansion prediction method and device
CN113421202A (en) * 2021-06-25 2021-09-21 浙江大华技术股份有限公司 Method and device for correcting low-frequency heterogeneity of infrared image and storage medium
CN117372285A (en) * 2023-12-05 2024-01-09 成都市晶林科技有限公司 Time domain high-pass filtering method and system for static and dynamic region distinction

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150319387A1 (en) * 2014-04-30 2015-11-05 Ulis Method of infrared image processing for non-uniformity correction
DE102015109878B3 (en) * 2015-06-19 2016-07-28 Basler Ag Correction of a patterned image defect
CN108665423A (en) * 2017-12-26 2018-10-16 西安电子科技大学 Infrared image non-uniformity correction method based on guiding filtering and high-pass filtering
CN109741267A (en) * 2018-12-05 2019-05-10 西安电子科技大学 Infrared Image Non-uniformity Correction method based on three sides filtering and neural network
CN110246108A (en) * 2018-11-21 2019-09-17 浙江大华技术股份有限公司 A kind of image processing method, device and computer readable storage medium
CN110390690A (en) * 2019-07-11 2019-10-29 Oppo广东移动通信有限公司 Depth map treating method and apparatus
CN110400273A (en) * 2019-07-11 2019-11-01 Oppo广东移动通信有限公司 Filtering method, device, electronic equipment and the readable storage medium storing program for executing of depth data

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150319387A1 (en) * 2014-04-30 2015-11-05 Ulis Method of infrared image processing for non-uniformity correction
DE102015109878B3 (en) * 2015-06-19 2016-07-28 Basler Ag Correction of a patterned image defect
CN108665423A (en) * 2017-12-26 2018-10-16 西安电子科技大学 Infrared image non-uniformity correction method based on guiding filtering and high-pass filtering
CN110246108A (en) * 2018-11-21 2019-09-17 浙江大华技术股份有限公司 A kind of image processing method, device and computer readable storage medium
CN109741267A (en) * 2018-12-05 2019-05-10 西安电子科技大学 Infrared Image Non-uniformity Correction method based on three sides filtering and neural network
CN110390690A (en) * 2019-07-11 2019-10-29 Oppo广东移动通信有限公司 Depth map treating method and apparatus
CN110400273A (en) * 2019-07-11 2019-11-01 Oppo广东移动通信有限公司 Filtering method, device, electronic equipment and the readable storage medium storing program for executing of depth data

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111968066A (en) * 2020-08-27 2020-11-20 烟台艾睿光电科技有限公司 Infrared image correction method, device and equipment and refrigeration infrared imaging system
CN111968066B (en) * 2020-08-27 2023-01-10 烟台艾睿光电科技有限公司 Infrared image correction method, device and equipment and refrigeration infrared imaging system
CN112651886A (en) * 2020-11-25 2021-04-13 杭州微帧信息科技有限公司 Method for removing color bands in enhanced image by combining original image
CN113349810A (en) * 2021-05-27 2021-09-07 北京安德医智科技有限公司 Cerebral hemorrhage focus identification and hematoma expansion prediction method and device
CN113349810B (en) * 2021-05-27 2022-03-01 北京安德医智科技有限公司 Cerebral hemorrhage focus identification and hematoma expansion prediction system and device
CN113421202A (en) * 2021-06-25 2021-09-21 浙江大华技术股份有限公司 Method and device for correcting low-frequency heterogeneity of infrared image and storage medium
CN117372285A (en) * 2023-12-05 2024-01-09 成都市晶林科技有限公司 Time domain high-pass filtering method and system for static and dynamic region distinction
CN117372285B (en) * 2023-12-05 2024-02-20 成都市晶林科技有限公司 Time domain high-pass filtering method and system for static and dynamic region distinction

Similar Documents

Publication Publication Date Title
CN111080561A (en) Time domain high-pass filtering method
Zuo et al. New temporal high-pass filter nonuniformity correction based on bilateral filter
CN110211056B (en) Self-adaptive infrared image de-striping algorithm based on local median histogram
US8810602B2 (en) Image processing apparatus, medium, and method
Wang et al. Enhancing low light videos by exploring high sensitivity camera noise
CN109272520B (en) Self-adaptive infrared focal plane non-uniform correction method combining motion guidance and edge detection
CN113313641A (en) CT image denoising method with self-adaptive median filtering
CN106197690B (en) Image calibrating method and system under the conditions of a kind of wide temperature range
CN109813442B (en) Multi-frame processing-based internal stray radiation non-uniformity correction method
Dharejo et al. A color enhancement scene estimation approach for single image haze removal
CN108965879B (en) Space-time domain self-adaptive just noticeable distortion measurement method
CN114092353A (en) Infrared image enhancement method based on weighted guided filtering
CN109934790A (en) Infrared imaging system asymmetric correction method with adaptive threshold
CN103076096A (en) Infrared nonuniformity correcting algorithm based on mid-value histogram balance
Li et al. Scene-based nonuniformity correction based on bilateral filter with reduced ghosting
CN112150386B (en) SAR image speckle non-local average inhibition method based on contrast mean value
CN110782403A (en) Infrared image non-uniformity correction method
CN110349113B (en) Adaptive image defogging method based on dark primary color priori improvement
CN105869129A (en) Residual heterogeneous noise elimination method for aiming at thermal infrared image after heterogeneous correction
Li et al. Interior radiation noise reduction method based on multiframe processing in infrared focal plane arrays imaging system
CN111932478A (en) Self-adaptive non-uniform correction method for uncooled infrared focal plane
CN111445435B (en) Multi-block wavelet transform-based reference-free image quality evaluation method
CN110009575B (en) Infrared image stripe noise suppression method based on sparse representation
CN114549386A (en) Multi-exposure image fusion method based on self-adaptive illumination consistency
CN110852977B (en) Image enhancement method for fusing edge gray level histogram and human eye visual perception characteristics

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