CN112967305A - Image cloud background detection method under complex sky scene - Google Patents

Image cloud background detection method under complex sky scene Download PDF

Info

Publication number
CN112967305A
CN112967305A CN202110312071.5A CN202110312071A CN112967305A CN 112967305 A CN112967305 A CN 112967305A CN 202110312071 A CN202110312071 A CN 202110312071A CN 112967305 A CN112967305 A CN 112967305A
Authority
CN
China
Prior art keywords
edge
image
neighborhood
extracting
contour
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.)
Granted
Application number
CN202110312071.5A
Other languages
Chinese (zh)
Other versions
CN112967305B (en
Inventor
朱伟
刘羽
董小舒
邱文嘉
辛付豪
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.)
Nanjing Laisi Electronic Equipment Co ltd
Original Assignee
Nanjing Laisi Electronic Equipment Co ltd
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 Nanjing Laisi Electronic Equipment Co ltd filed Critical Nanjing Laisi Electronic Equipment Co ltd
Priority to CN202110312071.5A priority Critical patent/CN112967305B/en
Publication of CN112967305A publication Critical patent/CN112967305A/en
Application granted granted Critical
Publication of CN112967305B publication Critical patent/CN112967305B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • 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
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/181—Segmentation; Edge detection involving edge growing; involving edge linking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • 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
    • G06T2207/10016—Video; Image sequence
    • 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
    • G06T2207/20024—Filtering details
    • 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
    • G06T2207/20036—Morphological image processing

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)

Abstract

The invention provides an image cloud background detection method under a complex sky scene, aiming at an empty application scene of a photoelectric search and tracking system, firstly constructing a neighborhood convolution kernel template, extracting first and second neighborhood envelope characteristics, and performing relevant neighborhood filtering on an original image; secondly, constructing multi-direction morphological gradient direction structural operators of horizontal, vertical, right diagonal and secondary diagonal, extracting contour edges of the image, and adaptively constructing a direction edge weight factor to fuse multi-direction detection results to obtain a contour edge image; and finally, carrying out neighborhood connected domain marking on the edge image to realize the extraction of the cloud background contour.

Description

Image cloud background detection method under complex sky scene
Technical Field
The invention relates to the field of image processing and computer vision, in particular to an image cloud background detection method under a complex sky scene.
Background
With the increasing complexity of modern war environment, the photoelectric searching and tracking system has more extensive requirements in various battlefield defense scenes due to the characteristics of passivity, difficulty in being influenced by the environment and the like. Aiming at the problem that a thick cloud layer shields a target in photoelectric space detection of a small target, the existing solution mainly comprises a neural network, morphological filtering, an interframe difference method and the like, which have good performance when the background changes smoothly, but the detection performance is greatly influenced when the frame frequency is low under the background of a complex cloud layer, and the algorithm has a high false alarm.
Disclosure of Invention
The purpose of the invention is as follows: the invention aims to solve the technical problem of the prior art, and provides a method for detecting an image cloud background under a complex sky scene, which comprises the following steps:
step 1, constructing a 5-by-5 neighborhood convolution kernel template S, extracting relevant first and second neighborhood envelope feature median values, and performing neighborhood convolution operation on an original Image1 to obtain a relevant neighborhood filtering Image 2;
step 2, constructing a directional gradient edge operator, and carrying out morphological directional gradient contour edge extraction on the Image 2;
step 3, extracting the detail edge of the outline to obtain a unidirectional edge detection result, and constructing a fusion edge weight factor lambdaiCalculating the edge detection in the horizontal, vertical, major diagonal and minor diagonal directions to obtain an edge Image 3;
and 4, marking the connected domain of the edge Image3, and extracting to obtain the cloud background connected domain outline.
The step 1 comprises the following steps:
step 1-1, defining a 5 x 5 neighborhood convolution kernel template as S, and setting an initial value of an S matrix as 0;
step 1-2, calculating a first neighborhood W of a window of a current image center f (x, y)1Median value m of1I.e. m1＝median(W1) Wherein mean is the median operation of the calculated vectors, x is the horizontal coordinate, y is the vertical coordinate, W1Comprises the following steps:
W1＝[f(x-1,y)，f(x,y+1)，f(x,y)，f(x+1,y)，f(x,y-1)]
step 1-3, calculating a second neighborhood W of a window of a current image center f (x, y)2Median value m of2I.e. m2＝median(W2)，W2Comprises the following steps:
Figure BDA0002990191500000021
step 1-4, setting convolution kernel template center pixel S (2,2) as Max (m)1,m2) And obtaining a neighborhood convolution kernel template S, wherein Max is the maximum operation of the calculation vector.
In step 2, the constructing a directional gradient edge operator includes:
respectively establishing gradient structural operators U in horizontal, vertical, main diagonal and sub diagonal directions1、U2、U3、U4Operator U of direction gradient structureiComprises the following steps:
Figure BDA0002990191500000022
Figure BDA0002990191500000023
wherein i is 1,2,3, 4.
In step 2, the performing morphological direction gradient contour edge extraction on the Image2 includes:
the following formula is adopted for extracting the morphological direction gradient contour edge:
Figure BDA0002990191500000024
ZIi＝f·Ui-fΘUi·L
wherein ZEiFor crude extraction of the results at the edge of the outer contour, ZIiThe results are crudely extracted for the inner contour edge, f is Image2,
Figure BDA0002990191500000025
for morphological dilation operation, Θ is a morphological erosion operation,
Figure BDA0002990191500000026
for a morphological open operation,. for a morphological close operation, the structure operator L is:
Figure BDA0002990191500000031
in step 3, the extracting the contour detail edge to obtain a one-way edge detection result includes:
and extracting the detail edge of the outline by adopting the following formula to obtain a one-way edge detection result:
ZXi＝Max(ZEi,ZIi)-Min(ZEi,ZIi)
Zi＝ZEi+ZIi+ZXi
wherein ZXiFor contour detail edges, ZiFor the one-directional edge detection result, Min represents the minimum value of the calculation vector.
In step 3, the fusion edge weight factor lambda is constructediThe method for detecting and calculating the edges in the horizontal, vertical, major diagonal and minor diagonal directions to obtain an edge Image3 includes:
constructing a fused edge weight factor lambda by the following formulai：
Di＝∑|di-d0|
Figure BDA0002990191500000032
Wherein DiRepresenting gradient gray scale difference in each direction, when i takes values of 1,2,3 and 4 respectively, diRespectively representing the pixel values of horizontal, vertical, major diagonal, and minor diagonal, d0Is the center pixel value;
extracting the multi-direction edge detection result to generate a final edge detection result Z, which is the edge Image 3:
Figure BDA0002990191500000033
step 4 comprises the following steps: traversing the edge Image3, extracting pixel points with the pixel value of 1 in the Image3 as seed pixel points, marking all foreground pixel points with the pixel values of 4 neighborhoods or 8 neighborhoods adjacent to the seed pixel points as L, and generating a connected domain outline Q by marking all pixel composition areas with the same mark as L1Repeating the above steps until the whole image is traversed to obtain the final connected domain vector (Q)1,Q2,...Qi),i∈(1,2,3...)，QiRepresenting the ith connected domain profile.
The invention mainly extracts the cloud layer and background edge characteristics by providing improved morphological gradient edge characteristics, and can effectively detect a cloud layer background area and pre-judge a shielded area of a target by fusing adaptive weighting through morphological gradients in all directions.
Has the advantages that: the invention provides an image cloud background detection method under a complex sky scene, aiming at the scene that the image has cloud interference to an empty detection target. By means of relevant neighborhood filtering, interference of image noise on target background edges is effectively reduced, outline characteristics of a cloud layer background are accurately described by utilizing a multidirectional morphological gradient operator, and generation and interference of false edges are greatly reduced.
Drawings
The foregoing and/or other advantages of the invention will become further apparent from the following detailed description of the invention when taken in conjunction with the accompanying drawings.
Fig. 1 is a flowchart of image cloud background detection in a complex sky scene according to an embodiment of the present invention.
FIG. 2 is a diagram of a first neighborhood in an embodiment of the present invention.
FIG. 3 is a diagram of a second neighborhood in an embodiment of the present invention.
FIG. 4 is a schematic diagram of the morphological directional gradient in an embodiment of the invention.
Fig. 5 is an input image in an embodiment of the present invention.
Fig. 6 is a result of image edge detection in the embodiment of the present invention.
Fig. 7 shows the contour detection result in the embodiment of the present invention.
Detailed Description
As shown in fig. 1, the present invention provides a method for detecting an image cloud background under a complex sky scene, including:
a) as shown in fig. 5, the infrared image data image1 is acquired, a 5 × 5 neighborhood convolution kernel template S is constructed, all values of the matrix are set to 0, and as shown in fig. 2 and 3, a 5 × 5 pixel matrix fw of the current image center f (x, y) is extracted1：
Figure BDA0002990191500000041
Calculating a first neighborhood W of the current window1Median value m (172,166,178,176,174)1Similarly, a second neighborhood W of the current window is calculated 1742Median value m of2170, the convolution kernel template center pixel S (2,2) is set to Max (m)1,m2) Obtaining S (2,2) 170, wherein Max is the maximum value operation of the calculation vector, and performing neighborhood convolution S operation on traversal of the original Image1 to obtain a related neighborhood filtered Image 2;
b) constructing directional gradient edge operators, and in order to further enhance the extraction of cloud background edges, as shown in fig. 4, respectively establishing horizontal, vertical, main diagonal and sub diagonal directional gradient structure operators U1、U2、U3、U4Operator U of direction gradient structurei(i ═ 1,2,3,4) is:
Figure BDA0002990191500000051
Figure BDA0002990191500000052
the morphological direction gradient contour edge extraction comprises the following steps:
Figure BDA0002990191500000053
ZIi＝f·Ui-fΘUi·L
wherein ZEiFor crude extraction of the results at the edge of the outer contour, ZIiFor the coarse extraction of the inner contour edge, f is Image2, UiI is a directional gradient edge operator and takes the values of 1,2,3 and 4,
Figure BDA0002990191500000054
for morphological dilation operation, Θ is a morphological erosion operation,
Figure BDA0002990191500000055
for a morphological open operation,. for a morphological close operation, the structure operator L is:
Figure BDA0002990191500000056
c) extracting the detail edge of the outline to obtain a one-way edge detection result Zi：
ZXi＝Max(ZEi,ZIi)-Min(ZEi,ZIi)
Zi＝ZEi+ZIi+ZXi
Wherein ZXiFor details of the profileEdge, ZiMax is the maximum value of the calculation vector and Min is the minimum value of the calculation vector for the unidirectional edge detection result. As shown in fig. 6, the multi-direction edge detection result is extracted to generate a final edge detection result Z, which is an edge Image3, and specifically:
Figure BDA0002990191500000061
wherein an edge weight factor lambda is fusediThe method comprises the following steps:
Di＝∑|di-d0|
Figure BDA0002990191500000062
wherein DiRepresenting gradient gray-scale differences in various directions, diRespective pixel values, d, representing horizontal, vertical, major diagonal, minor diagonal0The value of i is 1,2,3 and 4 for the pixel value of the central point.
d) Traversing the edge Image3, extracting pixel points with the pixel value of 1 in the Image3 as seed pixel points, marking all the foreground pixel points with the pixel values of 1 in the adjacent 4 neighborhoods or 8 neighborhoods with L, and marking the same mark as L to form a region by all pixels to generate a connected domain Q1Repeating the above steps until the whole image is traversed to obtain the final connected domain vector (Q)1,Q2,...Qi) I ∈ (1,2, 3.). The final result is shown in fig. 7.
The present invention provides a method for detecting an image cloud background under a complex sky scene, and a number of methods and approaches for implementing the technical solution are provided, the above description is only a preferred embodiment of the present invention, and it should be noted that, for a person skilled in the art, a number of improvements and embellishments may be made without departing from the principle of the present invention, and these improvements and embellishments should also be regarded as the protection scope of the present invention. All the components not specified in the present embodiment can be realized by the prior art.

Claims (7)

1. A method for detecting an image cloud background under a complex sky scene is characterized by comprising the following steps:
step 1, constructing a 5-by-5 neighborhood convolution kernel template S, extracting relevant first and second neighborhood envelope feature median values, and performing neighborhood convolution operation on an original Image1 to obtain a relevant neighborhood filtering Image 2;
step 2, constructing a directional gradient edge operator, and carrying out morphological directional gradient contour edge extraction on the Image 2;
step 3, extracting the detail edge of the outline to obtain a unidirectional edge detection result, and constructing a fusion edge weight factor lambdaiCalculating the edge detection in the horizontal, vertical, major diagonal and minor diagonal directions to obtain an edge Image 3;
and 4, marking the connected domain of the edge Image3, and extracting to obtain the cloud background connected domain outline.
2. The method of claim 1, wherein step 1 comprises:
step 1-1, defining a 5 x 5 neighborhood convolution kernel template as S, and setting an initial value of an S matrix as 0;
step 1-2, calculating a first neighborhood W of a window of a current image center f (x, y)1Median value m of1I.e. m1＝median(W1) Wherein mean is the median operation of the calculated vectors, x is the horizontal coordinate, y is the vertical coordinate, W1Comprises the following steps:
W1＝[f(x-1,y)，f(x,y+1)，f(x,y)，f(x+1,y)，f(x,y-1)]
step 1-3, calculating a second neighborhood W of a window of a current image center f (x, y)2Median value m of2I.e. m2＝median(W2)，W2Comprises the following steps:
Figure FDA0002990191490000011
step 1-4, setting convolution kernel template center pixel S (2,2) as Max (m)1,m2) To obtain a neighborhoodAnd (4) convolution kernel template S, wherein Max is the maximum operation of the calculation vector.
3. The method of claim 2, wherein in step 2, said constructing directional gradient edge operators comprises:
respectively establishing gradient structural operators U in horizontal, vertical, main diagonal and sub diagonal directions1、U2、U3、U4Operator U of direction gradient structureiComprises the following steps:
Figure FDA0002990191490000021
Figure FDA0002990191490000022
wherein i is 1,2,3, 4.
4. The method according to claim 3, wherein in step 2, the performing morphological direction gradient contour edge extraction on the Image2 comprises:
the following formula is adopted for extracting the morphological direction gradient contour edge:
Figure FDA0002990191490000023
ZIi＝f·Ui-fΘUi·L
wherein ZEiFor crude extraction of the results at the edge of the outer contour, ZIiThe results are crudely extracted for the inner contour edge, f is Image2,
Figure FDA0002990191490000024
for morphological dilation operation, Θ is a morphological erosion operation,
Figure FDA0002990191490000025
for a morphological open operation,. for a morphological close operation, the structure operator L is:
Figure FDA0002990191490000026
5. the method according to claim 4, wherein in step 3, the extracting the contour detail edge to obtain a one-directional edge detection result comprises:
and extracting the detail edge of the outline by adopting the following formula to obtain a one-way edge detection result:
ZXi＝Max(ZEi,ZIi)-Min(ZEi,ZIi)
Zi＝ZEi+ZIi+ZXi
wherein ZXiFor contour detail edges, ZiFor the one-directional edge detection result, Min represents the minimum value of the calculation vector.
6. The method according to claim 5, wherein in step 3, the fusion edge weight factor λ is constructediThe method for detecting and calculating the edges in the horizontal, vertical, major diagonal and minor diagonal directions to obtain an edge Image3 includes:
constructing a fused edge weight factor lambda by the following formulai：
Di＝∑|di-d0|
Figure FDA0002990191490000031
Wherein DiRepresenting gradient gray scale difference in each direction, when i takes values of 1,2,3 and 4 respectively, diRespectively representing the pixel values of horizontal, vertical, major diagonal, and minor diagonal, d0Is the center pixel value;
extracting the multi-direction edge detection result to generate a final edge detection result Z, which is the edge Image 3:
Figure FDA0002990191490000032
7. the method of claim 6, wherein step 4 comprises: traversing the edge Image3, extracting pixel points with the pixel value of 1 in the Image3 as seed pixel points, marking all foreground pixel points with the pixel values of 4 neighborhoods or 8 neighborhoods adjacent to the seed pixel points as L, and generating a connected domain outline Q by marking all pixel composition areas with the same mark as L1Repeating the above steps until the whole image is traversed to obtain the final connected domain vector (Q)1,Q2,...Qi),i∈(1,2,3...)，QiRepresenting the ith connected domain profile.
CN202110312071.5A 2021-03-24 2021-03-24 Image cloud background detection method under complex sky scene Active CN112967305B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110312071.5A CN112967305B (en) 2021-03-24 2021-03-24 Image cloud background detection method under complex sky scene

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110312071.5A CN112967305B (en) 2021-03-24 2021-03-24 Image cloud background detection method under complex sky scene

Publications (2)

Publication Number Publication Date
CN112967305A true CN112967305A (en) 2021-06-15
CN112967305B CN112967305B (en) 2023-10-13

Family

ID=76278239

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110312071.5A Active CN112967305B (en) 2021-03-24 2021-03-24 Image cloud background detection method under complex sky scene

Country Status (1)

Country Link
CN (1) CN112967305B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117853432A (en) * 2023-12-26 2024-04-09 北京长木谷医疗科技股份有限公司 Hybrid model-based osteoarthropathy identification method and device
CN117853932A (en) * 2024-03-05 2024-04-09 华中科技大学 Sea surface target detection method, detection platform and system based on photoelectric pod

Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105004737A (en) * 2015-07-14 2015-10-28 浙江大学 Self-adaption improved gradient information-based fruit surface defect detection method
CN105574855A (en) * 2015-12-10 2016-05-11 南京理工大学 Method for detecting infrared small targets under cloud background based on temperate filtering and false alarm rejection
CN105741281A (en) * 2016-01-28 2016-07-06 西安理工大学 Image edge detection method based on neighbourhood dispersion
US20160267675A1 (en) * 2014-06-23 2016-09-15 Boe Technology Group Co., Ltd. Image edge detection method and apparatus thereof, image target identification method and apparatus thereof
CN106327522A (en) * 2016-08-25 2017-01-11 上海航天控制技术研究所 Infrared small target detection method based on multi-direction morphological filtering complex cloud background
CN106778499A (en) * 2016-11-24 2017-05-31 江苏大学 A kind of method of quick positioning people's eye iris during iris capturing
CN107578418A (en) * 2017-09-08 2018-01-12 华中科技大学 A kind of indoor scene profile testing method of confluent colours and depth information
CN108109155A (en) * 2017-11-28 2018-06-01 东北林业大学 A kind of automatic threshold edge detection method based on improvement Canny
CN108280823A (en) * 2017-12-29 2018-07-13 南京邮电大学 The detection method and system of the weak edge faults of cable surface in a kind of industrial production
CN111507426A (en) * 2020-04-30 2020-08-07 中国电子科技集团公司第三十八研究所 No-reference image quality grading evaluation method and device based on visual fusion characteristics
CN111681262A (en) * 2020-05-08 2020-09-18 南京莱斯电子设备有限公司 Method for detecting infrared dim target under complex background based on neighborhood gradient
CN112184639A (en) * 2020-09-15 2021-01-05 佛山(华南)新材料研究院 Round hole detection method and device, electronic equipment and storage medium
WO2021012757A1 (en) * 2019-07-23 2021-01-28 南京莱斯电子设备有限公司 Real-time target detection and tracking method based on panoramic multichannel 4k video images

Patent Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160267675A1 (en) * 2014-06-23 2016-09-15 Boe Technology Group Co., Ltd. Image edge detection method and apparatus thereof, image target identification method and apparatus thereof
CN105004737A (en) * 2015-07-14 2015-10-28 浙江大学 Self-adaption improved gradient information-based fruit surface defect detection method
CN105574855A (en) * 2015-12-10 2016-05-11 南京理工大学 Method for detecting infrared small targets under cloud background based on temperate filtering and false alarm rejection
CN105741281A (en) * 2016-01-28 2016-07-06 西安理工大学 Image edge detection method based on neighbourhood dispersion
CN106327522A (en) * 2016-08-25 2017-01-11 上海航天控制技术研究所 Infrared small target detection method based on multi-direction morphological filtering complex cloud background
CN106778499A (en) * 2016-11-24 2017-05-31 江苏大学 A kind of method of quick positioning people's eye iris during iris capturing
CN107578418A (en) * 2017-09-08 2018-01-12 华中科技大学 A kind of indoor scene profile testing method of confluent colours and depth information
CN108109155A (en) * 2017-11-28 2018-06-01 东北林业大学 A kind of automatic threshold edge detection method based on improvement Canny
CN108280823A (en) * 2017-12-29 2018-07-13 南京邮电大学 The detection method and system of the weak edge faults of cable surface in a kind of industrial production
WO2021012757A1 (en) * 2019-07-23 2021-01-28 南京莱斯电子设备有限公司 Real-time target detection and tracking method based on panoramic multichannel 4k video images
CN111507426A (en) * 2020-04-30 2020-08-07 中国电子科技集团公司第三十八研究所 No-reference image quality grading evaluation method and device based on visual fusion characteristics
CN111681262A (en) * 2020-05-08 2020-09-18 南京莱斯电子设备有限公司 Method for detecting infrared dim target under complex background based on neighborhood gradient
CN112184639A (en) * 2020-09-15 2021-01-05 佛山(华南)新材料研究院 Round hole detection method and device, electronic equipment and storage medium

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
张路;张志勇;肖山竹;卢焕章;: "基于多向背景预测的红外弱小目标检测", 信号处理, no. 11 *
蒋洁琦;杨庚;刘沛东;钱晨;: "表面弱边缘瑕疵检测算法及应用", 计算机技术与发展, no. 05 *
金长江;师廷伟;: "红外弱小目标检测背景抑制算法研究", 中国测试, no. 04 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117853432A (en) * 2023-12-26 2024-04-09 北京长木谷医疗科技股份有限公司 Hybrid model-based osteoarthropathy identification method and device
CN117853432B (en) * 2023-12-26 2024-08-16 北京长木谷医疗科技股份有限公司 Hybrid model-based osteoarthropathy identification method and device
CN117853932A (en) * 2024-03-05 2024-04-09 华中科技大学 Sea surface target detection method, detection platform and system based on photoelectric pod
CN117853932B (en) * 2024-03-05 2024-05-14 华中科技大学 Sea surface target detection method, detection platform and system based on photoelectric pod

Also Published As

Publication number Publication date
CN112967305B (en) 2023-10-13

Similar Documents

Publication Publication Date Title
Xuan et al. An improved canny edge detection algorithm
Werlberger et al. Motion estimation with non-local total variation regularization
CN110555818B (en) Method and device for repairing cloud region of satellite image sequence
CN111160407B (en) Deep learning target detection method and system
CN110660065B (en) Infrared fault detection and identification algorithm
CN105184766B (en) A kind of level set image segmentation method of frequency domain boundary energy model
Cao et al. Infrared small target detection based on derivative dissimilarity measure
CN112967305B (en) Image cloud background detection method under complex sky scene
Xu et al. COCO-Net: A dual-supervised network with unified ROI-loss for low-resolution ship detection from optical satellite image sequences
CN110826575A (en) Underwater target identification method based on machine learning
CN111783834B (en) Heterogeneous image matching method based on joint graph spectrum feature analysis
CN112837331A (en) Fuzzy three-dimensional SAR image target extraction method based on self-adaptive morphological reconstruction
WO2022233252A1 (en) Image processing method and apparatus, and computer device and storage medium
KR102173955B1 (en) Apparatus of detecting feature point for tracking object and method thereof
Yang et al. Hierarchical joint bilateral filtering for depth post-processing
Bai Overview of image mosaic technology by computer vision and digital image processing
CN111008555B (en) Unmanned aerial vehicle image small and weak target enhancement extraction method
CN110490877B (en) Target segmentation method for binocular stereo image based on Graph Cuts
Tian et al. High confidence detection for moving target in aerial video
Nie et al. Object-level hyperspectral target detection based on spectral-spatial features integrated yolov4-tiny network
CN110532989B (en) Automatic detection method for offshore targets
CN117636172B (en) Target detection method and system for weak and small target of remote sensing image
Jiao et al. Cost-volume filtering-based stereo matching with improved matching cost and secondary refinement
Hu et al. High Precision Automatic Extraction of Cultural Relic Diseases Based on Improved Slic and AP Clustering
Li et al. Object detection from MMS imagery using deep learning for generation of road orthophotos

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
GR01 Patent grant
GR01 Patent grant