CN104537632B - Infrared image histogram enhancement method based on edge extracting - Google Patents
Infrared image histogram enhancement method based on edge extracting Download PDFInfo
- Publication number
- CN104537632B CN104537632B CN201410834788.6A CN201410834788A CN104537632B CN 104537632 B CN104537632 B CN 104537632B CN 201410834788 A CN201410834788 A CN 201410834788A CN 104537632 B CN104537632 B CN 104537632B
- Authority
- CN
- China
- Prior art keywords
- image
- fringe region
- edge
- histogram
- gray value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of infrared image histogram enhancement method based on edge extracting, comprise the following steps：The fringe region of original infrared image is extracted, to obtain fringe region image；The gray value being not zero in statistics fringe region image, to obtain the histogram and its accumulative histogram of fringe region；The untreated infrared image is strengthened with accumulative histogram according to the histogram of the fringe region image.The Enhancement Method enhancing contrast of the present invention is high, image clearly, good visual effect.
Description
Technical field
The present invention relates to a kind of Infrared Image Enhancement Algorithm, specifically, the present invention relates to a kind of based on edge extracting
The method of infrared histogram enhancement algorithm.
Background technology
Infrared imaging system strong antijamming capability, concealment is good, and air penetration capacity is strong, can adapt to a variety of particular fields
Close.But it is due to infrared detector self-characteristic, such as sensitivity is high, the big characteristic of dynamic range, and in complex work environment
Various noise jammings, cause infrared image to show high background, low-contrast feature.Image scene is embodied in whole
Very small part is only account in the dynamic range of infrared imaging system, picture contrast is poor, smudgy.Therefore need to original
Infrared image carries out image enhaucament, so as to improve the contrast of infrared image, improves the visual effect of image.
At present, the infrared image enhancing method of main flow has histogram enhancement and gray scale stretching Enhancement Method.But this two
Class enhancing algorithm is all to carry out image enhancement processing according to the grey-level statistics in entire image, and have ignored image and exist in itself
Characteristic in spatial domain.But, the background with close gray scale occupies very big one wherein again for infrared image in itself
Fraction, and detailed information and Weak target only account for small proportion, it is easy to cause in image enhaucament
During, image detail floods in the background so that the identification difficulty to Small object becomes big.
The content of the invention
It is an object of the invention to provide a kind of Enhancement Method of infrared image, method is extracted by edge detection operator
Image border region, and image enhaucament is carried out to infrared image according to the histograms of these fringe regions.It is processed by the invention
Picture contrast afterwards is high, and details is clear, and quality is good.
Technical scheme is as follows：
A kind of infrared image histogram enhancement method based on edge extracting.Comprise the following steps：(1) extract original infrared
Image F fringe region, to obtain fringe region image Fedge；(2) statistics fringe region image FedgeIn the gray scale that is not zero
Value, to obtain the histogram p and its accumulative histogram c of the fringe region image；
(3) the original infrared image F is increased with the histogram p and accumulative histogram c of the fringe region image
By force, enhanced infrared image F' is obtained.
The step (1), extracts untreated original infrared image F fringe region, to obtain fringe region image
Fedge, specifically include following sub-step：The edge of the infrared image F is detected by edge detection operator, then according to edge
The gray value of pixel be set to 1, non-edge pixels gray value be set to 0 principle generation bianry image Tedge；Travel through the two-value
Image TedgeOn all pixels, for the pixel on each coordinate, if in pixel or its neighborhood on the coordinate
There is the pixel that gray value is 1 in pixel, then the fringe region image FedgeGrey scale pixel value in upper same coordinate is set to institute
The gray value at infrared image F same coordinates is stated, otherwise the fringe region image FedgeGrey scale pixel value on the upper coordinate
It is set to 0, complete bianry image T of traversaledgePerform and fringe region image F is obtained after above-mentioned corresponding processingedge。
The step (2), statistics meter fringe region image FedgeIn the gray value that is not zero, to obtain the fringe region
Histogram p and its accumulative histogram c the step of, specifically include procedure below：Travel through the fringe region image Fedge, statistics
The number of times that gray value k occurs in the fringe region image, to obtain the histogram p of the fringe region image, the Nogata
The each single item of figure with p (k) represent, k=0,1 ..., M, M be original infrared image F maximum gray scale；Accumulative fringe region figure
The histogram of picture is to obtain the accumulative histogram c, and each single item of the accumulative histogram is represented with c (k), and sets c (0)=0,
Computing formula is：
The step (3), is entered with the histogram p and accumulative histogram c of fringe region image to the original infrared image F
Row enhancing, the step of obtaining enhanced infrared image F', including following sub-step：By the histogram of the fringe region image
The corresponding k values coordinate of all nonzero terms in p arranges structure sequence v in order, and specific method is：The traversal institute from k=1 to k=M
The histogram p of fringe region image items are stated, by all k values met corresponding to the item of p (k) ≠ 0 according to from small to large
Order rearranges sequence v, and the every of sequence v is represented with v (i), i=1,2,3 ..., N, N be the fringe region image
The quantity of nonzero term in histogram p, and expand sequence v and obtain sequence v', sequence v' items v'(j) represent, it is specific to expand
Mode is：
The gray value k of each pixel on original infrared image F is calculated, enhancing result is calculated, schemes after being strengthened
As the gray value w (k) of pixel at upper same position, circular is：For the gray scale of each pixel on original image
Value k, is searched in sequence v', determines that k size is in v' between which two adjacency, specific lookup method is traversal
Sequence v', finds T and causes as j=T, v'(j)≤k ＜ v'(j+1) set up, mark T1=v'(T), T2=v'(T+1), and according to
Formula is calculated as below and calculates the final enhanced result of Enhancement Method：
Wherein w (k) is enhancing result,Expression is rounded downwards.
The Enhancement Method of the present invention improves texture, edge, small mesh by counting the histogram in infrared image edge region
The ratio that the regions such as mark are occupied in statistics with histogram so that substantially, image clearly after enhancing is contrasted grain details enhancing effect
Degree is high, good visual effect.
Brief description of the drawings
Fig. 1 is the infrared image histogram enhancement method flow diagram of the invention based on edge extracting.
Fig. 2 is the refined flow chart of step of the present invention (1).
Fig. 3 is the refined flow chart of step of the present invention (3).
Fig. 4 shows the image handled without Enhancement Method of the present invention.
Fig. 5 shows the fringe region image after step of the present invention (1) processing.
Fig. 6 shows to handle obtained fringe region histogram by step of the present invention (2).
Fig. 7 shows the image after Enhancement Method of the present invention processing.
Embodiment
The present invention is further described with reference to the accompanying drawings and examples.As shown in Figure 1, the present invention is carried based on edge
The infrared histogram enhancement algorithm taken comprises the following steps：
(1) original infrared image F fringe region is extracted, to obtain fringe region image FedgeThe step of, specifically include
Following sub-step：
By edge detection operator, such as Canny operators, the edge of the infrared image F is detected, then according to edge pixel
Gray value be set to 1, non-edge pixels gray value be set to 0 principle generation bianry image Tedge；
Travel through the bianry image TedgeOn all pixels, for the pixel on each coordinate, if on the coordinate
Pixel or its neighborhood in pixel there is the pixel that gray value is 1, then the fringe region image FedgeUpper same coordinate
On grey scale pixel value be set to gray value at the infrared image F same coordinates, the otherwise fringe region image FedgeOn
Grey scale pixel value on the coordinate is set to 0, complete bianry image T of traversaledgeEdge is obtained after performing above-mentioned corresponding processing
Area image Fedge；
(2) statistics fringe region image FedgeIn the gray value that is not zero, with obtain the fringe region histogram p and
The step of its accumulative histogram c, specifically include following sub-step：
Travel through the fringe region image Fedge, the number of times that statistics gray value k occurs in the fringe region image, with
Obtain the histogram p of the fringe region image, the histogrammic each single item is represented with p (k), k=0,1 ..., M, M to be original
Infrared image F maximum gray scale；The histogram of accumulative fringe region image is to obtain the accumulative histogram c, and this is accumulative straight
The each single item of square figure is represented with c (k), and sets c (0)=0, and computing formula is：
(3) the original infrared image F is increased with the histogram p and accumulative histogram c of the fringe region image
By force, the step of obtaining enhanced infrared image F', including following sub-step：
The corresponding k values coordinate of all nonzero terms in the histogram p of the fringe region image is arranged into structure in order
Sequence v, specific method is：The histogram p of traversal fringe region image items, p is met by all from k=1 to k=M
(k) the k values corresponding to item ≠ 0 rearrange sequence v according to order from small to large, and the every of sequence v is represented with v (i), i
=1,2,3 ..., N, N be the quantity of nonzero term in the histogram p of the fringe region image, and expand sequence v and obtain sequence
V', sequence v' items v'(j) represent, specific extended mode is：
The gray value k of each pixel on original infrared image F is calculated, enhancing result is calculated, schemes after being strengthened
As the gray value w (k) of pixel at upper same position, circular is：For the ash of each pixel on original image
Angle value k, is searched in sequence v', determines that k size is in v' between which two adjacency, having specific lookup method is
For ergodic sequence v', find T and cause as j=T, v'(j)≤k ＜ v'(j+1) set up, mark T1=v'(T), T2=v'(T+
1), and according to formula is calculated as below the final enhanced result of Enhancement Method is calculated：
Wherein w (k) is enhancing result,Expression takes downwards
It is whole.
As shown in figure 4, its be sweep type infrared system output, resolution sizes be 1024 × 1280, based on demarcation
Nonuniformity correction after infrared image.The infrared image is without image enhancement processing, it can be seen that picture contrast is poor,
Image detail is difficult to differentiate.
Fig. 5 is rim detection in the infrared fringe region image obtained after step of the present invention (1) processing, step (1)
It is to be handled by threshold parameter value for 0.005 Canny edge detection operators.Fig. 6 is that edge area image is carried out
Its histogram that statistics with histogram is obtained.
Fig. 7 is to handle obtained infrared image by Enhancement Method of the present invention, it can be seen that the contrast of image is high, carefully
Section is clear, and quality is good.
Claims (1)
1. a kind of infrared image histogram enhancement method based on edge extracting, it is characterised in that comprise the following steps：
(1) original infrared image F fringe region is extracted, to obtain fringe region image Fedge；
(2) statistics fringe region image FedgeIn the gray value that is not zero, with obtain the fringe region image histogram p and
Its accumulative histogram c；
(3) the original infrared image F is strengthened with the histogram p and accumulative histogram c of the fringe region image, obtained
To enhanced infrared image F'；
The step (1), extracts original infrared image F fringe region, to obtain fringe region image FedgeThe step of, specifically
Including procedure below：
The edge of the original infrared image F is detected by edge detection operator, is then set to according to the gray value of edge pixel
1st, the gray value of non-edge pixels is set to 0 principle generation bianry image Tedge；
Travel through the bianry image TedgeOn all pixels, for the pixel on each coordinate, if the pixel on the coordinate
Or there is the pixel that gray value is 1, then the fringe region image F in the pixel in its neighborhoodedgePicture in upper same coordinate
Plain gray value is set to the gray value at the infrared image F same coordinates, otherwise the fringe region image FedgeThe upper coordinate
On grey scale pixel value be set to 0, complete bianry image T of traversaledgeFringe region figure is obtained after performing above-mentioned corresponding processing
As Fedge；
The step (2), statistics fringe region image FedgeIn the gray value that is not zero, to obtain the Nogata of the fringe region
The step of scheming p and its accumulative histogram c, specifically includes procedure below：
Travel through the fringe region image Fedge, gray value k is in the fringe region image F for statisticsedgeThe number of times of middle appearance, with
Obtain the histogram p of the fringe region image, the histogrammic each single item is represented with p (k), k=0,1 ..., M, M to be original
Infrared image F maximum gray scale；The histogram of accumulative fringe region image is to obtain the accumulative histogram c, and this is accumulative straight
The each single item of square figure is represented with c (k), and sets c (0)=0, and computing formula is：
The step (3), is entered with the histogram p and accumulative histogram c of the fringe region image to the original infrared image F
Row enhancing, the step of obtaining enhanced infrared image F', including procedure below：
By the fringe region image FedgeHistogram p in the corresponding k values coordinate of all nonzero terms arrange structure sequence in order
V is arranged, specific method is：The histogram p of traversal fringe region image items, p (k) is met by all from k=1 to k=M
K values corresponding to ≠ 0 item rearrange sequence v according to order from small to large, and the every of sequence v is represented with v (i), i=1,
2,3 ..., N, N be the quantity of nonzero term in the histogram p of the fringe region image, and expand sequence v and obtain sequence v', should
Sequence v' items v'(j) represent, specific extended mode is：
The gray value k of each pixel on original infrared image F is calculated, enhancing result is calculated, after being strengthened on image
The gray value w (k) of pixel at same position, circular is：For the gray value of each pixel on original image
K, is searched in sequence v', determines that k size is in v' between which two adjacency, specific lookup method is traversal sequence
V' is arranged, T is found and causes as j=T, v'(j)≤k ＜ v'(j+1) set up, mark T1=v'(T), T2=v'(T+1), and according to
Formula is calculated as below and calculates the final enhanced result of Enhancement Method：
Wherein w (k) is enhancing result,Expression is rounded downwards.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410834788.6A CN104537632B (en) | 2014-12-26 | 2014-12-26 | Infrared image histogram enhancement method based on edge extracting |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410834788.6A CN104537632B (en) | 2014-12-26 | 2014-12-26 | Infrared image histogram enhancement method based on edge extracting |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104537632A CN104537632A (en) | 2015-04-22 |
| CN104537632B true CN104537632B (en) | 2017-07-11 |
Family
ID=52853151
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410834788.6A Active CN104537632B (en) | 2014-12-26 | 2014-12-26 | Infrared image histogram enhancement method based on edge extracting |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104537632B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105957030B (en) * | 2016-04-26 | 2019-03-22 | 成都市晶林科技有限公司 | One kind being applied to the enhancing of thermal infrared imager image detail and noise suppressing method |
| CN106127713B (en) * | 2016-06-30 | 2019-02-26 | 山东中安科创光电科技有限公司 | Small and weak gas leakage target enhancing and Fast implementation |
| CN106204617B (en) * | 2016-07-21 | 2018-10-02 | 大连海事大学 | Adapting to image binarization method based on residual image histogram cyclic shift |
| CN112634273B (en) * | 2021-03-10 | 2021-08-13 | 四川大学 | Brain metastasis segmentation system based on deep neural network and construction method thereof |
| CN116681696B (en) * | 2023-07-27 | 2023-10-20 | 东莞雅达高精密塑胶模具有限公司 | Mold quality monitoring method for automatic production equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101101669A (en) * | 2007-07-13 | 2008-01-09 | 天津大学 | Image reinforcement method for self-adaptive regulation according to edge and brightness |
| CN101211459A (en) * | 2007-12-21 | 2008-07-02 | 上海广电集成电路有限公司 | Boundary point processed histogram balancing method |
| CN101957986A (en) * | 2010-10-11 | 2011-01-26 | 华中科技大学 | Method for realizing infrared image enhancement by controlling long and short integrals |
| CN102521813A (en) * | 2011-11-21 | 2012-06-27 | 华中科技大学 | Infrared image adaptive enhancement method based on dual-platform histogram |
-
2014
- 2014-12-26 CN CN201410834788.6A patent/CN104537632B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101101669A (en) * | 2007-07-13 | 2008-01-09 | 天津大学 | Image reinforcement method for self-adaptive regulation according to edge and brightness |
| CN101211459A (en) * | 2007-12-21 | 2008-07-02 | 上海广电集成电路有限公司 | Boundary point processed histogram balancing method |
| CN101957986A (en) * | 2010-10-11 | 2011-01-26 | 华中科技大学 | Method for realizing infrared image enhancement by controlling long and short integrals |
| CN102521813A (en) * | 2011-11-21 | 2012-06-27 | 华中科技大学 | Infrared image adaptive enhancement method based on dual-platform histogram |
Non-Patent Citations (4)
| Title |
|---|
| A simple and effective histogram equalization approach to image enhancement;H.D. Cheng ，X.J. Shi;《Digital Signal Processing》;20040331;第158-170页 * |
| 一种基于组合方法的红外图像增强方法;张志忠 等;《红外技术》;20091020;第31卷(第10期);第592页第1-2段 * |
| 基于梯度直方图变换增强红外图像的细节;赵文达 等;《光学精密工程》;20140715;第22卷(第7期);第1962-1968页 * |
| 结合边缘细节局部自适应增强的直方图均衡化;冷璐 等;《微电子学与计算机》;20100105;第38-41页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104537632A (en) | 2015-04-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104537632B (en) | Infrared image histogram enhancement method based on edge extracting | |
| Xuan et al. | An improved canny edge detection algorithm | |
| Rong et al. | An improved CANNY edge detection algorithm | |
| CN103077521B (en) | A kind of area-of-interest exacting method for video monitoring | |
| CN105513035B (en) | The detection method and its system of human body concealment article in a kind of passive millimeter wave image | |
| Fabijańska | Variance filter for edge detection and edge-based image segmentation | |
| CN104134200B (en) | Mobile scene image splicing method based on improved weighted fusion | |
| CN108898132B (en) | Terahertz image dangerous article identification method based on shape context description | |
| CN106204617B (en) | Adapting to image binarization method based on residual image histogram cyclic shift | |
| CN104966285B (en) | A kind of detection method of salient region | |
| CN101976436B (en) | Pixel-level multi-focus image fusion method based on correction of differential image | |
| JP2019134269A5 (en) | ||
| CN101127908A (en) | Video image motion processing method and implementation device with global feature classification | |
| CN110189347B (en) | Method and terminal for measuring volume of object | |
| CN108875504B (en) | Image detection method and image detection device based on neural network | |
| Han et al. | Research and implementation of an improved canny edge detection algorithm | |
| CN109886935A (en) | A kind of road face foreign matter detecting method based on deep learning | |
| CN110414385A (en) | A kind of method for detecting lane lines and system based on homography conversion and characteristic window | |
| US8396297B2 (en) | Supervised edge detection using fractal signatures | |
| EP2807631B1 (en) | Device and method for detecting a plant against a background | |
| JP4818285B2 (en) | Congestion retention detection system | |
| Wu et al. | Image Edge Detection Based on Sobel with Morphology | |
| CN103914840A (en) | Automatic human body contour extracting method for non-simple background | |
| Anjana et al. | Color image enhancement using edge based histogram equalization | |
| CN109448010B (en) | Automatic four-side continuous pattern generation method based on content features |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |