CN114022473A - Horizon detection method based on infrared image - Google Patents
Horizon detection method based on infrared image Download PDFInfo
- Publication number
- CN114022473A CN114022473A CN202111374988.4A CN202111374988A CN114022473A CN 114022473 A CN114022473 A CN 114022473A CN 202111374988 A CN202111374988 A CN 202111374988A CN 114022473 A CN114022473 A CN 114022473A
- Authority
- CN
- China
- Prior art keywords
- horizon
- image
- edge
- detection result
- detection
- 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
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 103
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 21
- 238000003708 edge detection Methods 0.000 claims abstract description 17
- 238000009499 grossing Methods 0.000 claims abstract description 10
- 238000012216 screening Methods 0.000 claims abstract description 3
- 238000000034 method Methods 0.000 claims description 37
- 230000002159 abnormal effect Effects 0.000 claims description 18
- 238000001914 filtration Methods 0.000 claims description 13
- 230000035772 mutation Effects 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 7
- 230000000694 effects Effects 0.000 abstract description 8
- 230000011218 segmentation Effects 0.000 description 13
- 238000003384 imaging method Methods 0.000 description 10
- 230000009466 transformation Effects 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 5
- 238000007781 pre-processing Methods 0.000 description 4
- 230000008439 repair process Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 230000000877 morphologic effect Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 239000002131 composite material Substances 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000004043 dyeing Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012512 characterization method Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000000407 epitaxy Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw 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
- G06T7/13—Edge detection
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a horizon detection method based on infrared images, which comprises the following steps: performing edge-preserving smoothing treatment on an infrared image to be detected by using an edge-preserving filter to obtain an output image; performing edge detection on the output image by using a Canny edge detection algorithm, and screening out a candidate horizon point set corresponding to each column; searching a boundary point in each candidate horizon point set, enabling the sky column region gray mean value taking the boundary point as a boundary to be larger than the ground object column region gray mean value, enabling the sum of the sky column region gray variance and the ground object column region gray variance to be minimum, taking the boundary point as a primary optimal horizon point of a corresponding column, and enabling the primary optimal horizon points of all columns to form a primary detection result of the horizon; and repairing the preliminary detection result of the horizon line to obtain a repaired horizon line detection result. The invention obtains good ground level line detection effect and can keep better robustness and real-time performance for complex conditions such as cloud layers existing in the sky or staggered ground object buildings.
Description
Technical Field
The invention relates to the technical field of horizon line detection, in particular to a horizon line detection method based on infrared images.
Background
The horizon detection technology is a key link in the field of infrared early warning and target search, and particularly when an infrared early warning system is used for long-distance detection on a head-up plane, an observed scene is a composite image containing sky and ground objects. Due to the fact that the sky background and the ground background are greatly different, the sky background is clean, the ground object background is more complex and changeable, the false alarm rate and the false detection rate are increased due to ground clutter, noise interference and the like, and the infrared target is challenged to be accurately detected and identified. In order to avoid the influence of ground clutter on the detection precision of an infrared target when an infrared early warning system is used for long-distance detection at a head-up plane, accurate horizon detection is necessary for a sky-ground composite image. Through effective detection on the horizon, the target search range can be reduced, the region of interest is divided, the influence of ground objects is eliminated, and powerful guarantee is provided for the precision and the efficiency of target detection of the infrared early warning system.
At present, the domestic research is less aiming at the ground level detection problem of the infrared image. Currently, most of the detection algorithms mainly used are based on sea-sky-line detection methods, which mainly include a Hough transform method, a region segmentation method, and the like.
For example, the invention patent CN109978869A provides a sea-sky-line detection method based on gray level co-occurrence matrix and Hough transformation. Carrying out image quality enhancement on the collected original visible light video monitoring image; converting the preprocessed image into a gray image by using a gray conversion formula after dyeing; calculating gray level co-occurrence matrixes under different angle values for the obtained gray level images; determining a possible existing area of the sea antenna according to the texture change rate; carrying out gray scale morphological corrosion and expansion operation on the region by adopting a mathematical morphological method; removing noise by using improved weighted Gaussian blur based on histogram optimization to obtain an edge image; and fitting to the finally detected sea-sky-line.
The invention patent CN108022214A discloses a horizon detection method suitable for unmanned aerial vehicle foggy day flight. And performing dark primary color defogging on the read RGB image to obtain a clear defogged image and a dark primary color image improved by using a soft matting method for feature extraction. And after dyeing, sequentially carrying out image segmentation, morphological processing and edge detection on the improved dark primary color image. And finally, carrying out linear detection by using Hough transformation, and carrying out least square method by using a linear detection result to accurately obtain horizon information.
The invention patent CN105644785A discloses an unmanned aerial vehicle landing method based on an optical flow method and horizon detection, firstly, image preprocessing is carried out on a video shot by a camera fixed at the bottom of the unmanned aerial vehicle in the flight process; then, carrying out line detection of Hough transformation on each image to acquire horizon information in the image; calculating the horizon information to obtain the flight attitude of the current unmanned aerial vehicle, and detecting the attitude information of the unmanned aerial vehicle by adopting an optical flow method; and finally, filtering the unmanned aerial vehicle attitude detected by the light stream method and the horizon line by combining an unmanned aerial vehicle motion model and adopting an extended Kalman filtering method, and selecting correct horizon line information to realize the autonomous landing process based on the unmanned aerial vehicle.
Yang-Min et al propose a horizon detection algorithm under low visibility (study on horizon detection algorithm under low visibility [ J ] computer engineering and design, 2012(01): 238-. The algorithm firstly carries out gray level transformation on an image to enhance contrast, defines the energy of each pixel according to two stages of attention and binding in the process of human visual system identification and the related theory of edge detection, finds out candidate horizon line demarcation points by using a dynamic programming method, and finally obtains a real horizon line through Hough transformation.
Beam epitaxy et al propose a method for detecting the horizon of an unmanned aerial vehicle based on sky segmentation (detection of the horizon of an unmanned aerial vehicle based on sky segmentation [ J ]. modern computer 2021,1:73-77), which performs color space conversion on an image to be detected; performing sky segmentation operation on the converted image; then, edge detection is carried out on the segmented image by using a Canny operator; and finally, identifying and detecting the horizon by using Hough transformation.
Sunpyuxin et al propose an infrared image horizon detection algorithm based on semantic segmentation (infrared image horizon detection algorithm research [ J ] based on semantic segmentation photoelectric technology application, 2020,35(6):55-57,78), the algorithm introduces a depth semantic segmentation model into a horizon detection task, and uses Deeplab-v3+ algorithm as a semantic segmentation algorithm to realize effective segmentation of the sky and the ground.
The method comprises the steps of firstly smoothing an infrared image, then detecting an image edge through a Sobel gradient operator, and searching the edge by using a Harr linear characteristic comparison mode to determine the positions of a sky region and a quasi-horizon; completing the initial detection of the horizon position by optimally extracting the boundary energy; and carrying out reasonability analysis and repairing on the horizon by using a relevant technology of numerical analysis.
However, the existing 'infrared image horizon detection method based on semantic segmentation' introduces a depth semantic segmentation model into a horizon detection task, has high calculation complexity of an algorithm, and often cannot meet the real-time requirement for an early warning system with high frame frequency or high resolution. The existing sky segmentation-based unmanned aerial vehicle horizon detection method is based on extraction of a color space and is not suitable for infrared images. Other existing horizon line detection methods have the problems of large calculation amount and limited precision.
In addition, most detection algorithms mainly rely on Hough transform method, region segmentation method and the like in consideration of sea-sky-line detection method. However, these methods are mainly directed to the imaging characteristics of the sea-sky, such as the unity of the sea-sky background, the straight line characteristics of the sea-sky boundary, and so on. The imaging characteristics of the sea-sky line of the infrared image and the sea-sky line of the visible image are different. The infrared image features have variable fluctuation, the texture of the features is more complex, and the horizon is not a simple straight line in a strict sense. Therefore, when the method is applied to horizon detection, the method has certain limitation, and the detection result is not satisfactory. Therefore, it is very important to provide a real-time detection method for the horizon of the infrared image.
Disclosure of Invention
In view of the above problems, the present invention provides a ground level detection method based on infrared images, which obtains a good ground level detection effect and can maintain good robustness and real-time performance even under complex conditions such as existence of cloud layers in the sky or staggering of ground buildings and houses.
In order to achieve the purpose, the invention adopts the following technical scheme:
a horizon detection method based on infrared images comprises the following steps:
step 1: performing edge-preserving smoothing treatment on an infrared image to be detected by using an edge-preserving filter to obtain an output image subjected to edge-preserving smoothing filtering;
step 2: performing edge detection on the output image by using a Canny edge detection algorithm, and screening out a candidate horizon point set corresponding to each column of the output image;
and step 3: searching a boundary point in each candidate horizon point set, enabling the sum of the sky column region gray variance and the ground feature column region gray variance taking the boundary point as a boundary to be minimum and the sky column region gray mean to be greater than the ground feature column region gray mean, taking the boundary point as a primary optimal horizon point of a corresponding column, and enabling the primary optimal horizon points of all the columns to form a primary horizon detection result;
and 4, step 4: and repairing the preliminary detection result of the horizon line to obtain a repaired horizon line detection result.
Firstly, performing edge-preserving smoothing treatment on an infrared image to be detected by using an edge-preserving filter; then, extracting the edge contour of the filtering image by using a Canny edge detection algorithm; considering that the horizon of the infrared image has edge characteristics, and the sky and ground regions bounded by the horizon have respective consistent imaging characteristics, establishing an energy function based on the gray level variance and the minimum of the sky and ground object sequence regions by taking the edge characteristics and the gray level mean of the sky and ground object sequence regions as constraint conditions by taking the image as a unit, thereby completing the initial detection of the horizon position; and finally, eliminating abnormal points in the horizon by using a numerical analysis method based on gradient information mutation, and reasonably repairing the horizon to obtain an accurate and complete horizon.
Compared with the prior art, the invention has the following beneficial effects:
(1) compared with the traditional sea-sky-line-based detection method, the method fully considers the imaging characteristic of the infrared image and the nonlinear edge characteristic of the horizon, and has good applicability to the horizon detection of the infrared image;
(2) compared with the existing infrared image horizon detection method, the method screens out the candidate horizon point set corresponding to each column based on the edge-preserving filter and the Canny edge detection algorithm, extracts the preliminary detection result of the horizon by taking the constraint conditions that the sum of the gray variance of the sky column region and the gray variance of the ground object column region is minimum and the gray mean value of the sky column region is greater than the gray mean value of the ground object column region, repairs the preliminary detection result of the horizon, finally obtains the repaired horizon detection result, obtains good horizon detection effect, and can keep good robustness and real-time performance even in the presence of complicated conditions such as cloud layers or staggered sky of ground object buildings.
Drawings
FIG. 1 is a flow chart of a method for detecting horizon based on infrared images according to the present invention;
FIG. 2 is a diagram illustrating an initial detection result of the horizon of an infrared image of a scene according to the present invention;
FIG. 3 is a repaired horizon line detection result obtained after the preliminary detection result of the horizon line shown in FIG. 2 is repaired;
fig. 4 shows the results of the horizon detection after partial patching of the infrared images of different scenes.
Detailed Description
The technical solution of the present invention will be described in detail with reference to the accompanying drawings and preferred embodiments.
The overall flow of the infrared image-based horizon detection method is shown in fig. 1, and the method comprises four steps, specifically comprising the following steps:
step 1: image pre-processing
In the step, an edge-preserving smoothing process is performed on the infrared image to be detected by using an edge-preserving filter, so that the complex texture structure of the ground object is removed, the edge is protected, the detection result is prevented from being influenced, and the output image after the edge-preserving smoothing filter is obtained after the processing.
The edge-preserving filter utilized in the image preprocessing of the step is rapid guide filtering
When the edge-preserving filter adopts the fast guide filtering, the image preprocessing process in the embodiment of the invention specifically comprises the following steps:
firstly, a Gaussian filter is utilized to remove small structures of an infrared image, and an output image G of the Gaussian filter is taken as a guide image of the next step, wherein the formula is as follows:
G＝Gussian(I，σr) (1)
wherein I is the input infrared image, σrThe size of the Gaussian kernel, namely the radius of the filtering window, is changed, and the filter effect of different degrees can be realized. And the image output by the Gaussian filtering is used as a guide image of the next step.
Considering the real-time performance of the algorithm, the invention uses the fast guide filtering to carry out the edge-preserving operation, and the formula is as follows:
Iout＝FastGuidedFilte r(G，I，σr，σs) (2)
wherein G is a guide diagram, σsFor regularizing parameters, IoutNamely the output image after the edge-preserving smooth filtering.
Step 2: edge detection
In general, the horizon of an infrared image has a boundary between the sky and a ground object, has obvious gradient change and conforms to edge characteristics, so that the output image obtained in the step 1 is subjected to edge detection by adopting a Canny edge detection algorithm in the step, and candidate horizons corresponding to each row of the output image are screened outA set of points. Taking the column of the output image as a unit, in the j-th column, the edge point detected by the Canny edge detection algorithm, namely the collection of candidate horizon points is marked as PCj。
And step 3: horizon point extraction
Consider that in an infrared image, the world region bounded by the optimal horizon has the following two imaging characteristics:
texture information of the ground object region is more complex than texture information of the sky region;
the sky area is brighter in imaging effect than the ground object area.
The invention thus summarizes the corresponding mathematical characterization consistent with the imaging characteristics above:
1) the sum of the gray variance of the sky area and the gray variance of the ground feature area is minimum;
2) the average value of the sky area gray scale is larger than the average value of the ground object area gray scale.
Therefore, the imaging characteristic of the infrared image and the edge characteristic of the horizon are taken as constraint conditions, the gray variance and the minimum function based on the upper-column region (namely sky column region) and the lower-column region (namely ground object column region) are established by taking the column as a unit, and the constraint condition that the average value of the gray of the sky column region is greater than the average value of the gray of the ground object column region is taken as a constraint condition, so that the initial detection of the position of the horizon is completed.
Specifically, a demarcation point P in a candidate horizon point set corresponding to each column is foundjThe sum of the sky column region gray variance and the ground object column region gray variance with the boundary point as the boundary point is minimized, and the requirement that the sky column region gray mean is larger than the ground object column region gray mean is met. Suppose that in column j, the demarcation point PjSatisfies the following conditions:
wherein, J (P)j) In the j-th column with PjThe energy value at the boundary point is the sum of the gray level variances of the upper and lower regions, IU(I, j) and ID(i, j) each represents PjThe gray scale values of the ith row and jth column pixels at the demarcation point, H and W respectively represent the height and width of the image,
and
respectively represent by PjThe gray average value of the sky column area and the ground object column area at the boundary point is calculated according to the following formula:
in the formula, NUAnd NDRespectively represent by PjThe number of pixels of the sky column region and the ground object column region at the boundary point.
The preliminary optimal horizon point of the j column can be solved according to the formulas (3) to (6)
The preliminary optimal horizon points of all the columns constitute a preliminary horizon detection result.
Still referring to fig. 1, step 3 specifically includes the following cyclic process:
step 31: dividing a jth column of the output image into a sky column region and a ground object column region by taking one candidate horizon point as a boundary point in the jth column of the output image;
step 32: then judging whether the sky column area gray mean value is larger than the ground matter column area gray mean value and the sum of the sky column area gray variance and the ground matter column area gray variance is minimum, if so, executing a step 33, otherwise, abandoning the candidate horizon line point and returning to the step 31, and reselecting a candidate horizon line point in a jth column as a boundary point;
step 33: taking the candidate horizon point as a primary optimal horizon point of a jth column;
step 34: and judging whether the jth column is the last column of the output image, if so, executing the step 4, and otherwise, returning to the step 31 after j + +.
And 4, step 4: and repairing the preliminary detection result of the horizon line to obtain a repaired horizon line detection result.
The general horizon can be detected through the steps 1 to 3, but the detection result often has some abnormal point sections, so the abnormal point sections need to be removed, and the horizon is reasonably repaired.
According to the regional characteristics of the abnormal point section of the statistical analysis of the preliminarily detected trend characteristics of the horizon, the invention provides a numerical analysis method based on gradient information mutation to eliminate the abnormal points in the preliminary detection result of the horizon so as to reasonably repair the horizon, and the method specifically comprises the following steps:
step 41: calculating an absolute gradient value of a preliminary detection result of the horizon;
the set of preliminary optimal horizon points for all columns of the output image is known as { P }1 best…Pj best，Pj+1 best…, the absolute gradient value of the horizon point corresponding to the jth column is:
the absolute gradient value of the preliminary detection result of the horizon line is obtained as G1，G2…Gj，Gj+1…}。
Step 42: searching abnormal point segments in the preliminary detection result of the horizon according to the absolute gradient value obtained by calculation in the step 41 and a preset gradient mutation threshold;
assume a predetermined gradient jump threshold value th, e.g. t in this embodimentAnd h is 10. If G ismIs the first gradient jump value, i.e. GmIf > th, the corresponding primary optimal horizon point P is consideredm bestIs a horizon break point. If G isnIs the next gradient jump value, i.e. GnIf m is more than th and m is more than n, the corresponding primary optimal horizon point P is consideredn bestAlso called horizon break points. If | m-n | ∈ [0, l), where l is the maximum length of the outlier segment, for example, l is 4 in this embodiment. Then the outlier segment is considered to be { P }m best…Pn bestAnd the abnormal point section is a point section consisting of the initial optimal horizon points corresponding to any two adjacent gradient mutation values and all the initial optimal horizon points between the two initial optimal horizon points. It should be noted that the values of the gradient mutation threshold and the maximum length of the outlier segment used in searching the outlier segment are not limited to 10 and 4 given in this embodiment, and other reasonable values may be adopted to achieve the effect of the present invention.
Step 43: and removing the abnormal point section from the preliminary detection result of the horizon, and replacing the abnormal point section by the previous preliminary optimal horizon point adjacent to the abnormal point section to obtain the repaired horizon detection result.
The outlier segment is known as { P }m best…Pn bestDivide the abnormal point into { P }m best…Pn bestRemoving the abnormal point section from the preliminary detection result of the horizon line, and reasonably repairing the horizon line by using the following formula:
{Pm best…Pn best}＝Pm-1 best (8)
and obtaining the final repaired horizon detection result after replacement.
Firstly, performing edge-preserving smoothing treatment on an infrared image to be detected by using an edge-preserving filter; then, extracting the edge contour of the filtering image by using a Canny edge detection algorithm; considering that the horizon of the infrared image has edge characteristics, and the sky and ground regions bounded by the horizon have respective consistent imaging characteristics, establishing an energy function based on the gray level variance and the minimum of the sky and ground object sequence regions by taking the edge characteristics and the gray level mean of the sky and ground object sequence regions as constraint conditions by taking the image as a unit, thereby completing the initial detection of the horizon position; and finally, eliminating abnormal points in the horizon by using a numerical analysis method based on gradient information mutation, and reasonably repairing the horizon to obtain an accurate and complete horizon.
Compared with the prior art, the invention has the following beneficial effects:
(1) compared with the traditional sea-sky-line-based detection method, the method fully considers the imaging characteristic of the infrared image and the nonlinear edge characteristic of the horizon, and has good applicability to the horizon detection of the infrared image;
(2) compared with the existing infrared image horizon detection method, the method screens out the candidate horizon point set corresponding to each column based on the edge-preserving filter and the Canny edge detection algorithm, extracts the preliminary detection result of the horizon by taking the constraint conditions that the sum of the gray variance of the sky column region and the gray variance of the ground object column region is minimum and the gray mean value of the sky column region is greater than the gray mean value of the ground object column region, repairs the preliminary detection result of the horizon, finally obtains the repaired horizon detection result, obtains good horizon detection effect, and can keep good robustness and real-time performance even in the presence of complicated conditions such as cloud layers or staggered sky of ground object buildings.
Fig. 2 shows the preliminary detection result of the horizon of the infrared image in a certain scene obtained through steps 1 to 3, and fig. 3 shows the repaired horizon detection result obtained by repairing the preliminary detection result of the horizon shown in fig. 2. As can be seen by comparing the graph 2 with the graph 3, through reasonable repair of the preliminary detection result of the horizon line, certain abnormal point sections are removed, so that the repaired horizon line detection result is closer to the real horizon line.
In order to verify the robustness and accuracy of the invention, the invention also performs the experiment of horizon detection on the infrared images of different scenes under complex conditions, the detection steps are as the steps 1 to 4, and partial horizon detection results are as shown in fig. 4. The detection result shows that the method has good ground level detection effect, and can keep better robustness and real-time performance even if cloud layers exist in the sky or complicated conditions such as ground object building staggering and the like.
The technical features of the embodiments described above may be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the embodiments described above are not described, but should be considered as being within the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (6)
1. A horizon detection method based on infrared images is characterized by comprising the following steps:
step 1: performing edge-preserving smoothing treatment on an infrared image to be detected by using an edge-preserving filter to obtain an output image subjected to edge-preserving smoothing filtering;
step 2: performing edge detection on the output image by using a Canny edge detection algorithm, and screening out a candidate horizon point set corresponding to each column of the output image;
and step 3: searching a boundary point in each candidate horizon point set, enabling the sum of the sky column region gray variance and the ground feature column region gray variance taking the boundary point as a boundary to be minimum and the sky column region gray mean to be greater than the ground feature column region gray mean, taking the boundary point as a primary optimal horizon point of a corresponding column, and enabling the primary optimal horizon points of all the columns to form a primary horizon detection result;
and 4, step 4: and repairing the preliminary detection result of the horizon line to obtain a repaired horizon line detection result.
2. The method for detecting the horizon based on the infrared image as claimed in claim 1, wherein the process of repairing the preliminary detection result of the horizon comprises the following steps:
step 41: calculating an absolute gradient value of the preliminary detection result of the horizon;
step 42: searching abnormal point sections in the preliminary detection result of the horizon according to the absolute gradient value and a preset gradient mutation threshold;
step 43: and removing the abnormal point section from the preliminary detection result of the horizon, and replacing the abnormal point section with the previous preliminary optimal horizon point adjacent to the abnormal point section to obtain the repaired horizon detection result.
3. The infrared-image-based horizon detecting method according to claim 2, characterized in that the outlier segment is a segment composed of preliminary optimal horizon points corresponding to any two adjacent gradient mutation values and all the preliminary optimal horizon points therebetween.
4. The method according to claim 2 or 3, wherein the gradient jump threshold is 10, and the maximum length of the outlier segment is 4.
5. The infrared image-based horizon detection method as claimed in claim 1, wherein the edge-preserving filter is a fast-guided filter.
6. The infrared image-based horizon detection method according to claim 5, wherein when the edge-preserving filter is a fast guided filter, step 1 comprises the following steps:
and removing small structures of the infrared image by using a Gaussian filter, and taking an image output by the Gaussian filter as a guide image of the next step, wherein the formula is as follows:
G＝Gussian(I,σr) (1)
where G is the guide image, I is the input infrared image, σrIs the gaussian kernel size;
and performing edge protection operation by using fast guided filtering, wherein the formula is as follows:
Iout＝FastGuidedFilte r(G,I,σr,σs) (2)
wherein σsFor regularizing parameters, IoutThe smooth filtered output image is edge-protected.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111374988.4A CN114022473B (en) | 2021-11-19 | 2021-11-19 | Horizon detection method based on infrared image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111374988.4A CN114022473B (en) | 2021-11-19 | 2021-11-19 | Horizon detection method based on infrared image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114022473A true CN114022473A (en) | 2022-02-08 |
| CN114022473B CN114022473B (en) | 2024-04-26 |
Family
ID=80065236
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111374988.4A Active CN114022473B (en) | 2021-11-19 | 2021-11-19 | Horizon detection method based on infrared image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114022473B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140293052A1 (en) * | 2011-07-08 | 2014-10-02 | Bendix Commercial Vehicle Systems Llc | Image-based vehicle detection and distance measuring method and apparatus |
| CN205160655U (en) * | 2012-12-21 | 2016-04-13 | 菲力尔系统公司 | A infrared imaging system for vehicle |
| CN109242870A (en) * | 2018-07-13 | 2019-01-18 | 上海大学 | A kind of sea horizon detection method divided based on image with textural characteristics |
| CN109493353A (en) * | 2018-10-06 | 2019-03-19 | 中国航空工业集团公司洛阳电光设备研究所 | A kind of sky and ocean background is plunged into the commercial sea the method for detecting in day line of demarcation |
| WO2020007307A1 (en) * | 2018-07-03 | 2020-01-09 | 深圳岚锋创视网络科技有限公司 | Sky filter method for panoramic images and portable terminal |
| CN113239733A (en) * | 2021-04-14 | 2021-08-10 | 重庆利龙科技产业(集团)有限公司 | Multi-lane line detection method |
-
2021
- 2021-11-19 CN CN202111374988.4A patent/CN114022473B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140293052A1 (en) * | 2011-07-08 | 2014-10-02 | Bendix Commercial Vehicle Systems Llc | Image-based vehicle detection and distance measuring method and apparatus |
| CN205160655U (en) * | 2012-12-21 | 2016-04-13 | 菲力尔系统公司 | A infrared imaging system for vehicle |
| WO2020007307A1 (en) * | 2018-07-03 | 2020-01-09 | 深圳岚锋创视网络科技有限公司 | Sky filter method for panoramic images and portable terminal |
| CN109242870A (en) * | 2018-07-13 | 2019-01-18 | 上海大学 | A kind of sea horizon detection method divided based on image with textural characteristics |
| CN109493353A (en) * | 2018-10-06 | 2019-03-19 | 中国航空工业集团公司洛阳电光设备研究所 | A kind of sky and ocean background is plunged into the commercial sea the method for detecting in day line of demarcation |
| CN113239733A (en) * | 2021-04-14 | 2021-08-10 | 重庆利龙科技产业(集团)有限公司 | Multi-lane line detection method |
Non-Patent Citations (3)
| Title |
|---|
| 宋敏敏;王爽;吕弢;袁瑜键;: "一种天地复杂背景下的红外弱小目标检测方法", 红外技术, no. 10, 18 October 2018 (2018-10-18) * |
| 毕国玲;赵建;隋龙;张艳超;: "基于红外图像的SF_6检漏定位系统", 光电工程, no. 03, 15 March 2011 (2011-03-15) * |
| 高海峰;张生伟;聂青凤;: "基于红外图像的天地线检测方法研究", 电光与控制, no. 07, 1 June 2016 (2016-06-01) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114022473B (en) | 2024-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107316031B (en) | Image feature extraction method for pedestrian re-identification | |
| CN108665487B (en) | Transformer substation operation object and target positioning method based on infrared and visible light fusion | |
| CN108230264B (en) | Single image defogging method based on ResNet neural network | |
| CN111104943B (en) | Color image region-of-interest extraction method based on decision-level fusion | |
| CN109242791B (en) | Batch repair method for damaged plant leaves | |
| CN110309781B (en) | House damage remote sensing identification method based on multi-scale spectrum texture self-adaptive fusion | |
| CN111046880A (en) | Infrared target image segmentation method and system, electronic device and storage medium | |
| CN111079556A (en) | Multi-temporal unmanned aerial vehicle video image change area detection and classification method | |
| CN111260616A (en) | Insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization | |
| CN109685045B (en) | Moving target video tracking method and system | |
| CN109978848B (en) | Method for detecting hard exudation in fundus image based on multi-light-source color constancy model | |
| JP2020085546A (en) | System for supporting inspection and repair of structure | |
| CN111160407A (en) | Deep learning target detection method and system | |
| Masood et al. | Plants disease segmentation using image processing | |
| CN109359554A (en) | A kind of forest fire recognition methods based on more composograph processing techniques | |
| CN113744191A (en) | Automatic cloud detection method for satellite remote sensing image | |
| CN113780110A (en) | Method and device for detecting weak and small targets in image sequence in real time | |
| CN111310771A (en) | Road image extraction method, device and equipment of remote sensing image and storage medium | |
| CN114881869A (en) | Inspection video image preprocessing method | |
| CN110264434B (en) | Single image rain removing method based on low-rank matrix completion | |
| CN114758139B (en) | Method for detecting accumulated water in foundation pit | |
| Khan et al. | Shadow removal from digital images using multi-channel binarization and shadow matting | |
| CN114022473B (en) | Horizon detection method based on infrared image | |
| CN114119634B (en) | Building automatic extraction method and system combining vegetation rejection and image feature consistency constraint | |
| CN113689399B (en) | Remote sensing image processing method and system for power grid identification |
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 |