CN110765631B - Effective imaging pixel-based small target judgment method for infrared radiation characteristic measurement - Google Patents
Effective imaging pixel-based small target judgment method for infrared radiation characteristic measurement Download PDFInfo
- Publication number
- CN110765631B CN110765631B CN201911052223.1A CN201911052223A CN110765631B CN 110765631 B CN110765631 B CN 110765631B CN 201911052223 A CN201911052223 A CN 201911052223A CN 110765631 B CN110765631 B CN 110765631B
- Authority
- CN
- China
- Prior art keywords
- target
- image
- effective imaging
- pixel
- imaging pixel
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/20—Drawing from basic elements, e.g. lines or circles
- G06T11/203—Drawing of straight lines or curves
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Algebra (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Photometry And Measurement Of Optical Pulse Characteristics (AREA)
- Transforming Light Signals Into Electric Signals (AREA)
Abstract
The invention discloses a small target judgment method based on infrared radiation characteristic measurement of effective imaging pixels, which comprises the steps of establishing a discrete system imaging model based on a point spread function, wherein a series of square target images with different theoretical side lengths are generated based on the established discrete system imaging model in a simulation mode; acquiring the number of target effective imaging pixels in an image; calculating the ratio of the target theoretical imaging pixel number to the effective imaging pixel number to obtain a characteristic curve of the ratio changing along with the target theoretical side length; and taking the number of target effective imaging pixels when the characteristic curve rises to a set value as the number of threshold effective imaging pixels for small target judgment, judging the target with the number of effective imaging pixels smaller than the number of threshold effective imaging pixels as the small target, and otherwise, judging the target as the surface target. The method has the characteristics of strong operability and wide applicability, and can be used as a universal judgment method.
Description
Technical Field
The invention relates to a small target judgment method based on effective imaging pixel infrared radiation characteristic measurement.
Background
The infrared radiation characteristic measurement can acquire information of target such as radiation brightness distribution, radiation intensity, temperature and the like, and is important basic data for evaluating the target detection and identification capability and the target infrared stealth capability of an optical system. Generally, the targets are divided into surface targets and small targets according to the number of pixels imaged on the target surface of the detector by the targets. The energy radiated by the surface object to the space is related to the observation visual direction angle and the specific part of the radiation from the object, and the radiation distribution characteristic of the surface object is described by the radiation brightness. For small targets, the relation between radiation energy and observation visual angle is mainly focused, and the radiation characteristics of the small targets are described by using radiation intensity. The imaging process of the small target is influenced by factors such as atmospheric disturbance, diffraction of an optical system, down-sampling of a sensor and the like, the energy of the imaged target is dispersed to a plurality of pixels, the dispersed image energy distribution area is larger than an ideal geometric imaging area, the edge of the target image is fuzzy, and the peak gray level is reduced. If a small object is treated as an area object, describing the radiance characteristics of the object using radiance will cause large errors. Therefore, there is a need for effectively distinguishing whether an object is a surface object or a small object in infrared radiation characteristic measurement.
Different research institutions give different plane target and small target distinguishing references, and some targets with the ideal geometric imaging pixel number of the target projection area smaller than 10 multiplied by 10pixel or 12 multiplied by 12pixel are defined as small targets; some targets have a ratio of the angle of view of the target to the optical imager to the instantaneous field of view of the optical imager of greater than 10, and point targets have a ratio of less than (equal to) 1. The essence of the method is based on the ideal geometric imaging pixel number of the target projection area on the infrared focal plane. When the infrared radiation characteristic of the target is actually measured, particularly when the radiation characteristic of the aerial flight target is measured, the actual external dimension of the target is difficult to obtain, and the posture of the target relative to the measuring equipment changes in real time, so that the ideal geometric imaging pixel number of the target to the measuring equipment is difficult to obtain. It can be seen that when measuring the infrared radiation intensity of a small target at present, there are two problems: firstly, the number of ideal geometric imaging pixels of a target relative to the projection area of a measuring device in actual measurement is difficult to obtain, and the current small target judgment method based on the number of the ideal geometric imaging pixels of the target is difficult to apply in practice; secondly, a unified judging method is lacked for distinguishing the surface target and the small target, the difference of the imaging effect of the infrared imaging systems of different models on the target is larger, the current method taking a certain determined ideal imaging pixel number as a judging reference does not fully consider the imaging characteristics of the infrared imaging systems of different models, and the applicability is not strong.
Disclosure of Invention
The invention aims to provide a small target judgment method based on effective imaging pixel infrared radiation characteristic measurement.
In order to achieve the purpose, the technical scheme of the invention is as follows:
a small target judgment method based on effective imaging pixel infrared radiation characteristic measurement is characterized in that a discrete system imaging model based on a point spread function is established, wherein a series of square target images with different theoretical side lengths are generated based on the established discrete system imaging model in a simulation mode; acquiring the number of target effective imaging pixels in an image; calculating the ratio of the number of target theoretical imaging pixels to the number of effective imaging pixels to obtain a characteristic curve of the ratio changing along with the target theoretical side length; and taking the number of target effective imaging pixels when the characteristic curve rises to a set value as the number of threshold effective imaging pixels for judging the small target, and judging the target with the number of effective imaging pixels smaller than the number of threshold effective imaging pixels as the small target, otherwise, judging the target as the surface target.
The scheme is further as follows: the set value is 0.8.
The scheme is further as follows: the step of generating a series of square target images with different theoretical side lengths based on the discrete system imaging model simulation comprises the following steps:
step one, establishing a discrete system imaging model based on a point spread function:
firstly, the method comprises the following steps: a discrete system point spread function is determined,
performing an inverse fourier transform on the optical transfer function of the infrared imaging system represented by equation 1 yields the point spread function of the continuous system represented by equation 2,
O tot (u,v)＝D(u,v)×O opt (u, v) formula 1
P c (X,Y)＝FF -1 {D(u,v)×O opt (u, v) } equation 2
Wherein:
O opt (u, v) limiting the contribution to the optical transfer function for the optical system;
d (u, v) is the contribution of the focal plane array constraints to the optical transfer function;
u and v are the spatial frequencies in the horizontal and vertical directions, respectively;
FF -1 representing a two-dimensional inverse fourier transform;
sampling the continuous system point spread function to obtain a discrete system point spread function under the high-resolution grid represented by formula 3,
wherein:
n 1 and n 2 Pixel indices in horizontal and vertical directions, respectively;
T 1 and T 2 Respectively the pixel center distances of the infrared system imaging focal plane array in the horizontal and vertical directions;
L 1 and L 2 Are respectively corresponding to T 1 And T 2 A proportionality constant of (a);
II, secondly: establishing a discrete system imaging model;
firstly discretizing a continuous image at a certain sampling interval to obtain a discretized image under a high-resolution grid, convolving the discretized image with a discrete system point spread function under the high-resolution grid corresponding to a system to obtain a degraded image under the high-resolution grid represented by a formula 4, downsampling the degraded image under the high-resolution grid and overlapping noise to obtain a discrete image output by the system represented by a formula 5;
wherein:
f c (x, y) are consecutive real images;
* Represents a convolution;
s represents a discrete process of continuous images, the sampling intervals of the discrete process are respectively
And
η(n 1 ,n 2 ) For noise, the mean is set to 0;
step two, generating a series of square target images with different theoretical side lengths through simulation:
generating a square target image with the side length of i pixels on a high-resolution grid, obtaining a discrete image output by a system generated by simulation based on the discrete system imaging model, wherein the target side length theoretical imaging pixel number and the target area theoretical imaging pixel number in the image are respectively represented by a formula 6,
wherein:
k and M r Respectively outputting the target side length theoretical imaging pixel number and the target area theoretical imaging pixel number in the system output image generated by simulation;
and step three, repeating the step two, changing the value i to enable the change range of the K value to be 1-100, and simulating to generate a series of target images with different K values.
The scheme is further as follows: the number of target effective imaging pixels in the acquired image is:
making the gray level in the image larger than the background gray level mean value
And 3 times background gray standard deviation sigma b The pixel of the sum is defined as a target effective imaging pixel, that is, the gray threshold of the effective imaging pixel is expressed by equation 7:
when the gray scale of the pixel in the target image
Then, the pixel is the effective imaging pixel.
The scheme is further as follows: the method further comprises the steps of suppressing noise and improving the statistical accuracy of effective imaging pixels, and the process is as follows:
1) Setting a gray threshold value in the image, and binarizing the simulation generated image according to the gray threshold value to obtain a binarized target image;
2) Performing digital morphological expansion operation on the binarized target image to obtain an expanded binarized target image, counting connected regions in the image, and recording a pixel index range corresponding to the connected region with the largest area as A;
3) Extracting a simulation generated image g d (n 1 ,n 2 ) Obtaining an image g of an effective imaging pixel distribution area by using the pixels belonging to the area A deff (n 1 ,n 2 ) Comparing the effective imaging pixel distribution area image with a gray threshold value to obtain
A corresponding binary matrix;
4) Calculating the sum of all elements in the binary matrix to obtain the effective imaging pixel number M eff 。
The invention has the beneficial effects that: a set of simulation analysis method is established, and a characteristic curve which changes along with the theoretical side length of the simulation target by the ratio of the number of target theoretical imaging pixels to the number of effective imaging pixels is obtained by introducing the concept of target effective imaging pixels. The characteristic curve rises as the theoretical edge length of the target increases. When the theoretical side length of the target is smaller, the characteristic curve rapidly rises along with the increase of the theoretical side length; after the characteristic curve rises to 0.8, the rising speed is reduced, and the curve tends to be flat. Thus, the corresponding target effective imaging pixel number when the characteristic curve is equal to 0.8 is taken as the threshold effective imaging pixel number for small target determination. And judging the target with the effective imaging pixel number smaller than the threshold effective imaging pixel number as a small target, otherwise, judging the target as an area target. The method solves the problem that the current judgment method based on the target theoretical imaging pixel number is difficult to use in actual measurement. Meanwhile, the method is based on a simulation analysis mode, considers the influence of a point spread function of an imaging optical system, has the characteristics of strong operability and wide applicability, and can be used as a universal judgment method.
The invention is described in detail below with reference to the figures and examples.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a block diagram of a discrete system imaging model;
FIG. 3 is a flow chart of a method for counting the number of effective imaging pixels;
fig. 4 is a typical characteristic graph.
Detailed Description
A small target judgment method based on effective imaging pixel infrared radiation characteristic measurement is characterized in that a discrete system imaging model based on a point spread function is established, wherein a series of square target images with different theoretical side lengths are generated based on the established discrete system imaging model in a simulation mode; acquiring the number of target effective imaging pixels in an image; calculating the ratio of the target theoretical imaging pixel number to the effective imaging pixel number to obtain a characteristic curve of the ratio changing along with the target theoretical side length; and taking the number of target effective imaging pixels when the characteristic curve rises to a set value as the number of threshold effective imaging pixels for judging the small target, and judging the target with the number of effective imaging pixels smaller than the number of threshold effective imaging pixels as the small target, otherwise, judging the target as the surface target.
The specific implementation steps are illustrated in fig. 1, wherein: as shown in fig. 4, the characteristic curve rises as the theoretical edge length of the target increases. When the theoretical side length of the target is smaller, the characteristic curve rapidly rises along with the increase of the theoretical side length; after the characteristic curve rises to 0.8-0.85, the rising speed is reduced, and the curve tends to be flat. Thus, the target effective imaging pixel number corresponding to the characteristic curve equal to 0.8 is taken as the threshold effective imaging pixel number for small target determination. And judging the target with the effective imaging pixel number smaller than the threshold effective imaging pixel number as a small target, otherwise, judging the target as an area target. The method solves the problem that the current judgment method based on the target theoretical imaging pixel number is difficult to use in actual measurement. Therefore, the set value is set to 0.8.
In the examples: the step of generating a series of square target images with different theoretical side lengths based on the discrete system imaging model simulation comprises the following steps:
step one, establishing a discrete system imaging model based on a point spread function:
firstly, the following steps: a discrete system point spread function is determined,
performing an inverse fourier transform on the optical transfer function of the infrared imaging system represented by equation 1 yields the point spread function of the continuous system represented by equation 2,
O tot (u,v)＝D(u,v)×O opt (u, v) formula 1
P c (X,Y)＝FF -1 {D(u,v)×O opt (u, v) } equation 2
Wherein:
O opt (u, v) limiting the contribution to the optical transfer function for the optical system;
d (u, v) is the contribution of the focal plane array constraints to the optical transfer function;
u and v are the spatial frequencies in the horizontal and vertical directions, respectively;
FF -1 representing a two-dimensional inverse fourier transform;
sampling the continuous system point spread function to obtain a discrete system point spread function under the high-resolution grid represented by formula 3,
wherein:
n 1 and n 2 Pixel indices in horizontal and vertical directions, respectively;
T 1 and T 2 Respectively the pixel center distances of the infrared system imaging focal plane array in the horizontal and vertical directions;
L 1 and L 2 Are respectively corresponding to T 1 And T 2 A proportionality constant of (a);
II, secondly, the method comprises the following steps: establishing a discrete system imaging model;
firstly discretizing a continuous image at a certain sampling interval to obtain a discretized image under a high-resolution grid, convolving the discretized image with a discrete system point spread function under the high-resolution grid corresponding to a system to obtain a degraded image under the high-resolution grid represented by a formula 4, downsampling the degraded image under the high-resolution grid and overlapping noise to obtain a discrete image output by the system represented by a formula 5;
wherein:
f c (x, y) are consecutive real images;
* Represents a convolution;
s represents a discrete process of continuous images, the sampling intervals of the discrete process are respectively
And
η(n 1 ,n 2 ) For noise, the mean is set to 0;
step two, generating a series of square target images with different theoretical side lengths through simulation:
generating a square target image with the side length of i pixels on a high-resolution grid, obtaining a discrete image output by a system generated by simulation based on the discrete system imaging model, wherein the number of target side length theoretical imaging pixels and the number of target area theoretical imaging pixels in the image are respectively represented by a formula 6,
wherein:
k and M r System outputs generated separately for simulationThe theoretical imaging pixel number of the target side length and the theoretical imaging pixel number of the target area in the image are obtained;
and step three, repeating the step two, changing the value i to enable the change range of the K value to be 1-100, and simulating to generate a series of target images with different K values. Here, the point spread function and the discrete system imaging model are not changed after the first step determination, so that here, only step two needs to be repeated and i is changed to generate simulated images of different K values.
Wherein: the number of target effective imaging pixels in the acquired image is:
making the gray level in the image larger than the background gray level mean value
And 3 times background gray standard deviation sigma b The pixel of the sum is defined as a target effective imaging pixel, that is, the gray threshold value of the effective imaging pixel is expressed by equation 7:
when the gray scale of the pixel in the target image
Then, the pixel is the effective imaging pixel.
The method further comprises the steps of suppressing noise and improving the statistical accuracy of effective imaging pixels, and the process is as follows:
1) Setting a gray threshold value in the image, and binarizing the simulation generated image according to the gray threshold value to obtain a binarized target image;
2) Performing digital morphological expansion operation on the binarized target image to obtain an expanded binarized target image, counting a connected region in the image, and recording a pixel index range corresponding to the connected region with the largest area as A;
3) Extracting a simulation generated image g d (n 1 ,n 2 ) Obtaining an image g of an effective imaging pixel distribution area by using the pixels belonging to the area A deff (n 1 ,n 2 ) Comparing the effective imaging pixel distribution area image with a gray threshold value to obtain
A corresponding binarization matrix;
4) Calculating the sum of all elements in the binary matrix to obtain the effective imaging pixel number M eff 。
The following is a detailed description taken in conjunction with the above examples:
(1) Establishing discrete system imaging model based on point spread function
1) Discrete system point spread function model
The optical transfer function of an infrared imaging system can be expressed as:
O tot (u,v)＝D(u,v)×O opt (u,v) (1)
wherein, O opt (u, v) limits the contribution to the optical transfer function for the optical system, and D (u, v) limits the contribution to the optical transfer function for the focal plane array, u and v being the spatial frequencies in the horizontal and vertical directions, respectively.
The optical transfer function induced by the optical system limitations mainly includes diffraction-limited contributions and aberration contributions. Among them, diffraction confinement plays a major role. For a circular aperture optical system, the optical transfer function due to diffraction limitations is:
in the formula (I), the compound is shown in the specification,
ρ 0 to the cut-off frequency:
in the formula, F/# is the F number of the optical system, and λ is the central wavelength of the response of the optical system.
The optical transfer function caused by the focal plane array mainly comprises an integral optical transfer function related to the size of a photosensitive element of the detector, a sampling optical transfer function related to the center distance of the image elements and a diffusion optical transfer function related to crosstalk between the image elements. For focal plane arrays with a high fill factor, the optical transfer function is mainly determined by the photosensor size, resulting in a point spread function of:
where a and b are the horizontal and vertical dimensions of the focal plane array photosensor, respectively. And d (x, y) is subjected to two-dimensional Fourier transform to obtain an optical transfer function caused by the focal plane array as follows:
substituting the formula (2) and the formula (5) into the formula (1) to obtain the total optical transfer function of the system, and performing inverse Fourier transform on the total optical transfer function to obtain the point spread function of the continuous system:
P c (X,Y)＝FF -1 {D(u,v)×O opt (u,v)} (6)
wherein, FF -1 Representing a two-dimensional inverse fourier transform.
Sampling the continuous system point spread function to obtain a discrete system point spread function under a high-resolution grid:
wherein n is 1 And n 2 Pixel indices, T, in the horizontal and vertical directions, respectively 1 And T 2 Respectively the pixel center distance of the infrared system imaging focal plane array in the horizontal and vertical directions,
and
horizontal and vertical sampling frequencies, respectively. L is 1 And L 2 Should be such that the sampling frequency is higher than 2 times the cut-off frequency of the optical transfer function of the system, i.e.:
for the convenience of simulation analysis, L is generally taken under the condition that equation (8) is satisfied 1 ＝L 2 。
2) Discrete system imaging model
As shown in fig. 2, for a continuous real image, discretizing the continuous image at a certain sampling interval to obtain a discretized image under a high-resolution grid, and convolving the discretized image with a discrete system point spread function under the high-resolution grid corresponding to the system to obtain a degraded image under the high-resolution grid:
wherein f is c (x, y) is a continuous real image, which represents convolution, S represents a discrete process of the continuous image, and the sampling intervals of the discrete process are respectively
And
downsampling the degraded image under the high-resolution grid and superposing noise to obtain a system output image:
wherein eta (n) 1 ,n 2 ) The average value is generally set to 0, and the standard deviation is equal to gaussian white noise of the gray standard deviation of the actually measured background image of the system.
g d (n 1 ,n 2 ) I.e. the discrete image output by the system as desired.
(2) Simulated image generation
Generating an ideal square target image with the side length of i pixels on a high-resolution grid, and finally obtaining a discrete image output by a system generated by simulation based on a discrete system imaging model. The number of target side length theoretical imaging pixels and the number of target area theoretical imaging pixels in the image are respectively as follows:
wherein, K and M r And respectively outputting the target side length theoretical imaging pixel number and the target area theoretical imaging pixel number in the system output image generated by the simulation.
By changing the value of i, a series of target images with different K values are generated in a simulation mode. Typically, K varies from 1 to 100.
(3) Target effective imaging pixel extraction method based on 3 sigma principle
And judging whether the pixels in the simulation generated image are target effective imaging pixels or not according to a 3 sigma principle. Making the gray level in the image larger than the background gray level mean value
And 3 times background gray standard deviation sigma b The summed pixels are defined as target effective imaging pixels, that is, the gray threshold of the effective imaging pixels is:
when target image pixel gray scale
Then, the pixel is the effective imaging pixel.
The flow of the statistical method for the number of effective imaging pixels is shown in fig. 3. In order to suppress the noise influence and improve the statistical accuracy, the specific statistical method is as follows:
1) Binarizing the simulation generated image by utilizing a gray threshold value to obtain a binarized target image;
2) Performing digital morphological expansion operation on the binarized target image to obtain an expanded binarized target image, counting a connected region in the image, and recording a pixel index range corresponding to the connected region with the largest area as A;
3) Extracting a simulation generated image g d (n 1 ,n 2 ) The pixels belonging to the area A are obtained to obtain an effective imaging pixel distribution area image g deff (n 1 ,n 2 ) Comparing the effective imaging pixel distribution area image with a gray threshold value to obtain
A corresponding binary matrix;
4) Calculating the sum of all elements in the binary matrix to obtain the effective imaging pixel number M eff 。
(4) Small target judgment method based on characteristic curve
Defining the ratio of the theoretical imaging pixel number of the target area to the effective imaging pixel number as follows:
because of the dispersion effect of the system point spread function, the target effective imaging pixel number obtained by the 3 σ principle is always larger than the theoretical imaging pixel number, and therefore R is a number smaller than 1. The larger the R is, the closer the target theoretical imaging pixel number and the effective imaging pixel number are, and the closer the target image gray distribution and the theoretical imaging gray distribution are, the target can be treated as a surface target; the smaller the R is, the larger the difference between the number of target theoretical imaging pixels and the number of effective imaging pixels is, the more target effective imaging pixels are caused by the dispersion effect in the image, the larger the difference between the target image gray distribution and the theoretical imaging gray distribution is, and the target is treated as a small target.
And counting the target effective imaging pixel number of a series of target images with different K values generated by simulation to obtain a characteristic curve of R changing along with K. A typical characteristic is shown in fig. 4. When the R value is smaller, the characteristic curve rapidly rises along with the increase of K; when R increases to 0.8, the rising speed of the characteristic curve slows down, and the curve becomes gentle. Threshold effective imaging pixel number M determined with the target effective imaging pixel number when R =0.8 as the small target cri The judgment criteria are established as follows:
whether the target in the actual image obtained by the infrared radiation characteristic measurement system is a small target can be judged by using the formula.
Details not described in the examples are known to the person skilled in the art.
Claims (3)
1. The method for judging the small target based on the infrared radiation characteristic measurement of the effective imaging pixel is characterized in that a series of square target images with different theoretical side lengths are generated based on the simulation of the established discrete system imaging model; acquiring the number of target effective imaging pixels in an image; calculating the ratio of the target theoretical imaging pixel number to the effective imaging pixel number to obtain a characteristic curve of the ratio changing along with the target theoretical side length; taking the number of target effective imaging pixels when the characteristic curve rises to a set value as the number of threshold effective imaging pixels for small target judgment, judging the target with the number of effective imaging pixels smaller than the number of threshold effective imaging pixels as a small target, and otherwise, judging the target as a surface target;
the step of generating a series of square target images with different theoretical side lengths based on the discrete system imaging model simulation comprises the following steps:
step one, establishing a discrete system imaging model based on a point spread function:
firstly, the method comprises the following steps: a discrete system point spread function is determined,
performing an inverse fourier transform on the optical transfer function of the infrared imaging system represented by equation 1 yields the point spread function of the continuous system represented by equation 2,
O tot (u,v)＝D(u,v)×O opt (u, v) formula 1
P c (X,Y)＝FF -1 {D(u,v)×O opt (u, v) } formula 2;
wherein:
O opt (u, v) limiting the contribution to the optical transfer function for the optical system;
d (u, v) is the contribution of the focal plane array constraints to the optical transfer function;
u and v are the spatial frequencies in the horizontal and vertical directions, respectively;
FF -1 representing a two-dimensional inverse fourier transform;
sampling the continuous system point spread function yields a discrete system point spread function under the high resolution grid represented by equation 3,
wherein:
n 1 and n 2 Pixel indices in horizontal and vertical directions, respectively;
T 1 and T 2 Respectively the pixel center distances of the infrared system imaging focal plane array in the horizontal and vertical directions;
L 1 and L 2 Are respectively corresponding to T 1 And T 2 A proportionality constant of (a);
II, secondly: establishing a discrete system imaging model;
for continuous real images, firstly discretizing the continuous images at certain sampling intervals to obtain discretized images under a high-resolution grid, convolving the discretized images with discrete system point spread functions under the high-resolution grid corresponding to a system to obtain degraded images under the high-resolution grid represented by a formula 4, and downsampling and superposing the degraded images under the high-resolution grid with noise to obtain discrete images output by the system represented by a formula 5;
wherein:
f c (x, y) are consecutive real images;
* Represents a convolution;
s represents a discrete process of continuous images, the sampling intervals of the discrete process are respectively
And
η(n 1 ,n 2 ) For noise, the mean is set to 0;
step two, generating a series of square target images with different theoretical side lengths through simulation:
generating a square target image with the side length of i pixels on a high-resolution grid, obtaining a discrete image output by a system generated by simulation based on the discrete system imaging model, wherein the target side length theoretical imaging pixel number and the target area theoretical imaging pixel number in the image are respectively represented by a formula 6,
wherein:
k and M r Respectively outputting the target side length theoretical imaging pixel number and the target area theoretical imaging pixel number in the system output image generated by simulation;
step three, repeating the step two, changing the value i to enable the change range of the K value to be 1-100, and simulating to generate a series of target images with different K values;
the number of target effective imaging pixels in the acquired image is:
making the gray level in the image larger than the background gray level mean value
And 3 times background gray standard deviation sigma b The pixel of the sum is defined as a target effective imaging pixel, that is, the gray threshold value of the effective imaging pixel is expressed by equation 7:
when the gray scale of the pixel in the target image
Then, the pixel is the effective imaging pixel.
2. The small object determination method according to claim 1, wherein the set value is 0.8.
3. The small target determination method according to claim 1, further comprising suppressing noise and improving the statistical accuracy of effective imaging pixels by:
1) Setting a gray threshold value in the image, and binarizing the simulation generated image according to the gray threshold value to obtain a binarized target image;
2) Performing digital morphological expansion operation on the binarized target image to obtain an expanded binarized target image, counting connected regions in the image, and recording a pixel index range corresponding to the connected region with the largest area as A;
3) Extracting a simulation generated image g d (n 1 ,n 2 ) The pixels belonging to the area A are obtained to obtain an effective imaging pixel distribution area image g deff (n 1 ,n 2 ) Comparing the effective imaging pixel distribution area image with a gray threshold value to obtain
A corresponding binarization matrix;
4) Calculating the sum of all elements in the binary matrix to obtain the effective imaging pixel number M eff 。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911052223.1A CN110765631B (en) | 2019-10-31 | 2019-10-31 | Effective imaging pixel-based small target judgment method for infrared radiation characteristic measurement |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911052223.1A CN110765631B (en) | 2019-10-31 | 2019-10-31 | Effective imaging pixel-based small target judgment method for infrared radiation characteristic measurement |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110765631A CN110765631A (en) | 2020-02-07 |
| CN110765631B true CN110765631B (en) | 2023-03-14 |
Family
ID=69334935
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911052223.1A Active CN110765631B (en) | 2019-10-31 | 2019-10-31 | Effective imaging pixel-based small target judgment method for infrared radiation characteristic measurement |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110765631B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111414800B (en) * | 2020-02-17 | 2023-08-01 | 妙微(杭州)科技有限公司 | Small target identification monitoring method in image and training set acquisition method thereof |
| CN111612710B (en) * | 2020-05-14 | 2022-10-04 | 中国人民解放军95859部队 | Geometric imaging pixel number calculation method for target rectangular projection image |
| CN114441595B (en) * | 2022-02-09 | 2022-07-29 | 四川省安全科学技术研究院 | Detection method for coal seam outcrop spontaneous combustion and influence range thereof |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014174438A1 (en) * | 2013-04-21 | 2014-10-30 | Nova Measuring Instruments Ltd. | Method and system for improving optical measurements on small targets |
| CN104268844A (en) * | 2014-10-17 | 2015-01-07 | 中国科学院武汉物理与数学研究所 | Small target infrared image processing method based on weighing local image entropy |
| CN104899866A (en) * | 2015-05-05 | 2015-09-09 | 河南三联网络技术有限公司 | Intelligent infrared small target detection method |
| CN105373657A (en) * | 2015-11-03 | 2016-03-02 | 广州市雷腾照明科技有限公司 | Design method for reflection cup forming rectangular spots |
| CN106845024A (en) * | 2017-03-02 | 2017-06-13 | 哈尔滨工业大学 | A kind of in-orbit imaging simulation method of optical satellite based on wavefront inverting |
| CN109523575A (en) * | 2018-11-12 | 2019-03-26 | 南通理工学院 | Infrared weak and small target detection method |
| CN109712158A (en) * | 2018-11-23 | 2019-05-03 | 山东航天电子技术研究所 | A kind of infrared small target catching method based on target background pixel statistical restraint |
-
2019
- 2019-10-31 CN CN201911052223.1A patent/CN110765631B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014174438A1 (en) * | 2013-04-21 | 2014-10-30 | Nova Measuring Instruments Ltd. | Method and system for improving optical measurements on small targets |
| CN104268844A (en) * | 2014-10-17 | 2015-01-07 | 中国科学院武汉物理与数学研究所 | Small target infrared image processing method based on weighing local image entropy |
| CN104899866A (en) * | 2015-05-05 | 2015-09-09 | 河南三联网络技术有限公司 | Intelligent infrared small target detection method |
| CN105373657A (en) * | 2015-11-03 | 2016-03-02 | 广州市雷腾照明科技有限公司 | Design method for reflection cup forming rectangular spots |
| CN106845024A (en) * | 2017-03-02 | 2017-06-13 | 哈尔滨工业大学 | A kind of in-orbit imaging simulation method of optical satellite based on wavefront inverting |
| CN109523575A (en) * | 2018-11-12 | 2019-03-26 | 南通理工学院 | Infrared weak and small target detection method |
| CN109712158A (en) * | 2018-11-23 | 2019-05-03 | 山东航天电子技术研究所 | A kind of infrared small target catching method based on target background pixel statistical restraint |
Non-Patent Citations (1)
| Title |
|---|
| 尺度空间理论在红外小目标检测中的应用研究;王铎;《光电技术应用》(第02期);77-80 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110765631A (en) | 2020-02-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110765631B (en) | Effective imaging pixel-based small target judgment method for infrared radiation characteristic measurement | |
| CN110287932B (en) | Road blocking information extraction method based on deep learning image semantic segmentation | |
| CN111028327B (en) | Processing method, device and equipment for three-dimensional point cloud | |
| Yang et al. | An automated cloud detection method based on the green channel of total-sky visible images | |
| Zhang et al. | Cloth simulation-based construction of pit-free canopy height models from airborne LiDAR data | |
| CN110969656B (en) | Detection method based on laser beam spot size of airborne equipment | |
| CN112561940B (en) | Dense multi-target parameter extraction method and device and terminal equipment | |
| CN110988909A (en) | TLS-based vegetation coverage determination method for sandy land vegetation in alpine and fragile areas | |
| WO2019215819A1 (en) | Synthetic aperture radar image analysis system, synthetic aperture radar image analysis method, and synthetic aperture radar image analysis program | |
| CN110288050B (en) | Hyperspectral and LiDar image automatic registration method based on clustering and optical flow method | |
| Bandyopadhyay et al. | Classification and extraction of trees and buildings from urban scenes using discrete return LiDAR and aerial color imagery | |
| JP2022013152A (en) | Cloud height measuring device, measuring point determination method, and cloud type determination method | |
| CN114299318A (en) | Method and system for rapid point cloud data processing and target image matching | |
| CN115457001A (en) | Photovoltaic panel foreign matter detection method, system, device and medium based on VGG network | |
| CN113781478B (en) | Oil tank image detection method, oil tank image detection device, electronic equipment and computer readable medium | |
| CN110365404B (en) | Wavefront-free sensing self-adaptive system and method for improving convergence speed by using same | |
| Xia et al. | Quality assessment for remote sensing images: approaches and applications | |
| CN108318458B (en) | Method for measuring outdoor typical feature pBRDF (binary RDF) suitable for different weather conditions | |
| Lyu et al. | Improvement of radar rainfall estimation using radar reflectivity data from the hybrid lowest elevation angles | |
| CN117671498A (en) | Unmanned aerial vehicle automatic detection system for lightning protection | |
| Liu et al. | Displacement field reconstruction in landslide physical modeling by using a terrain laser scanner–Part 1: Methodology, error analysis and validation | |
| Xiang et al. | Very short‐term prediction model for photovoltaic power based on improving the total sky cloud image recognition | |
| CN116883664A (en) | Remote sensing image segmentation method, device, equipment and computer readable storage medium | |
| KR20110048829A (en) | Apparatus and method for noise reduction of range images | |
| CN116612430A (en) | Method for estimating water level by utilizing video monitoring system based on deep learning |
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 |