JP5176289B2 - Infrared target selection device - Google Patents

Description

  The present invention relates to an infrared target selection device that selects a target from an infrared image. More specifically, the present invention relates to an infrared target selection device that is used in an infrared guidance device that guides a flying object toward a target and selects a target such as a ship from an infrared image.

  2. Description of the Related Art Conventionally, there has been known an infrared guidance device that detects infrared light from a target, selects a target, and guides a flying object toward the selected target (see, for example, Patent Documents 1 and 2).

Japanese Patent Laid-Open No. 2003-84054 (page 5, FIG. 1) JP 62-259112 A (page 3, Fig. 1)

Conventionally, when a target such as a ship floating on the sea is extracted from an infrared image captured by infrared rays, a point where the temperature is equal to or higher than a predetermined temperature (luminance) is used by utilizing the fact that there is generally a temperature difference between the target and the background. As a result, image processing for extracting a target has been performed. However, when the background is partially heated by sunlight or the like and the temperature difference between the target and the background becomes small, the infrared image captured by the infrared imaging device cannot take the contrast between the target and the background. However, there is a problem that the target cannot be recognized correctly, and as a result, the wrong object is selected.
In addition, when the target has a temperature distribution, the target shape shown in the infrared captured image is different from the actual shape, so that there is a problem that an incorrect target object is selected as a result of the image matching.

  The present invention has been made to solve such a problem, and even when the temperature difference between the target and the background is very small, it is possible to reliably select the correct target in the infrared captured image. It aims at obtaining the infrared target selection device which can be done.

In order to solve the above-described problems, an infrared target selection device according to the present invention separates infrared rays incident from the target and the background of the target into plane polarized light orthogonal to each other, and each of the separated infrared rays is arranged in two-dimensional coordinates. A difference between the electric signal amounts at the same coordinate position of the two-dimensional coordinates is obtained between the respective image signals received by an image sensor including a light receiving element, converted into an electric signal amount by the light receiving element, and output. A polarization image pickup device that outputs the obtained polarization difference image signal and outputs the luminance image signal obtained by adding the electric signal amount at the same coordinate position of the two-dimensional coordinate; and a target to be extracted; Preliminarily store data in which the combination of the background and the threshold value that enables separation of the target polarization difference image signal and the background polarization difference image signal in the combination are associated with each other A threshold database, be extracted to have targeted based on a combination of the background, binarization image obtained by binarizing the electric signal of the polarization difference image signal based on a comparison of the thresholds selected from the threshold database A binarization circuit that outputs a signal, and a value after binarization of the binarized image signal at the same coordinate position of the two-dimensional coordinates multiplied by the electric signal amount of the luminance image signal A target candidate extraction unit that extracts the target candidates from the obtained image.

  According to the infrared target selection device of the present invention, even if the temperature difference between the target and the background is very small, the target can be correctly and reliably selected in the infrared captured image.

Embodiment 1 FIG.
An infrared target selection apparatus according to Embodiment 1 of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram illustrating a configuration of an infrared target selection device 1 according to the first embodiment.

  In FIG. 1, an infrared target selection device 1 receives an infrared input light 100 from the outside, performs photoelectric conversion, and outputs a polarization difference analog image data 101 and luminance analog image data 106, and a polarization For each combination of the target and the background, an AD conversion circuit 11 that performs A / D (Analog / Digital) conversion on the differential analog image data 101 and outputs a polarization difference digital image signal 102, and a polarization difference digital image signal corresponding to the background. And a threshold value database 17 in which a threshold value that can be separated from the polarization difference digital image signal corresponding to the target is set in advance and stored, and the polarization difference digital image signal 102 is binarized using the threshold value to obtain a binarized digital signal. The binarization circuit 13 that outputs the image signal 103 and the luminance analog image data 106 are A / D converted to obtain the luminance digital image signal 1. 7, the binarized digital image signal 103 and the luminance digital image signal 107 are input and image processing is performed for each pixel of each image to extract information on the shape and luminance of the target candidate. The region feature quantity measurement unit 14 that outputs the target candidate extracted image 104, the target shape image database 20 in which target information 108 indicating the target shape and luminance is stored in advance, the target candidate extracted image 104, and the target information 108 And a target determination processing unit 15 that selects a target from the target candidate extraction images 104 by collation.

FIG. 2 is a diagram illustrating the configuration and operation of the polarization image capturing apparatus 10. The polarization image capturing apparatus 10 includes a polarizer 51, a difference signal processing unit 52, an addition signal processing unit 53, a first infrared image sensor 54, and a second infrared image sensor 55.
A polarizer 51 typified by an interference filter using a polarizing prism or a birefringent material has a P wave infrared ray 90P and an S wave infrared ray 90S, which are plane polarized lights orthogonal to each other from the infrared input light 100 incident from the target or its background. To output.
The first infrared image sensor 54 in which a plurality of light receiving elements are two-dimensionally arranged photoelectrically converts P-wave infrared light 90P and outputs a P-wave polarized image signal 100P after being converted into an electrical signal. Here, the P-wave polarized image signal 100P is information on the coordinate position (x, y) of the pixel corresponding to each of the two-dimensional array of light receiving elements constituting the first infrared image sensor 54, and the electrical signal at that coordinate position. Contains quantity data.
Similarly, the second infrared image sensor 55 in which a plurality of light receiving elements are two-dimensionally arranged performs photoelectric conversion of the S wave infrared light 90S and outputs an S wave polarized image signal 100S after conversion into an electrical signal. Similarly, the S-wave polarized image signal 100S includes information on the coordinate position (x, y) of the image pixel corresponding to each of the light receiving elements of the two-dimensional array constituting the second infrared image sensor 55, and the coordinate position at that coordinate position. It has electrical signal data.
Further, the coordinate axes and the origin positions of the first infrared image sensor 54 and the second infrared image sensor 55 are made to coincide with each other in advance.

Next, the addition signal processing unit 53 inputs the P wave polarization image signal 100P and the S wave polarization image signal 100S. The addition signal processing unit 53 uses the input P-wave polarization image signal 100P and the S-wave polarization image signal 100S to input the P-wave polarization image signals 100P and S from the light receiving elements at the same coordinate position (x, y). The process of adding the electric signal amount of the wave polarization image signal 100S is performed for each pixel unit and output as luminance analog image data 106.
The luminance analog image data 106 includes information on the coordinate position (x, y) of the pixel and data on the electric signal amount after the addition processing at the coordinate position.

  As described above, the addition signal processing unit 53 performs a calculation process of adding the electrical signal amounts of the P-wave polarization image signal 100P and the S-wave polarization image signal 100S obtained by being once separated by the polarizer 51 for each pixel. Thereby, luminance analog image data 106 is obtained as luminance information of the image of the infrared input light 100 incident from the target and the background thereof.

The differential signal processing unit 52 receives the P-wave polarization image signal 100P and the S-wave polarization image signal 100S. The differential signal processing unit 52 uses the input P-wave polarization image signal 100P and the S-wave polarization image signal 100S to input the P-wave polarization image signals 100P and S from the light receiving elements at the same coordinate position (x, y). A process for obtaining the difference in the electrical signal amount of the wave polarization image signal 100S is performed for each pixel unit, and output as polarization difference analog image data 101.
The polarization difference analog image data 101 includes information on the coordinate position (x, y) of a pixel and data after difference calculation processing at the coordinate position.

  As described above, the difference signal processing unit 52 performs a calculation process for obtaining the difference for each pixel in the electric signal amount of the P-wave polarized image signal 100P and the S-wave polarized image signal 100S obtained by being once separated by the polarizer 51. . By taking the difference, polarization difference analog image data 101 can be obtained as information on the degree of polarization corresponding to the image position of the infrared input light 100 incident from the target and its background.

  Here, the degree of polarization is, for example, a value represented by Equation 1 and indicates the degree of polarization of light rays. For example, the degree of polarization P is zero when there is no polarization of the light beam. As described above, the difference signal processing unit 52 obtains a difference between the P-wave polarization image signal 100P and the S-wave polarization image signal 100S, thereby obtaining a numerical value corresponding to the degree of polarization P shown in Expression (1).

  Here, the polarization characteristics (polarization degree) of electromagnetic waves radiated from an object are generally different due to differences in physical characteristics such as the material and surface state. For example, infrared rays are considered as electromagnetic waves, the target object is a direct infrared source, and the background is the sea surface that reflects sunlight. In such a case, the infrared rays from the target object do not have a specific polarization plane, whereas the reflected infrared rays from the background are mostly polarized in a certain direction. For this reason, in the example of FIG. 2, the reflected infrared light incident from the sea surface is attenuated by the polarizer 51, so that the contrast with a target having no specific polarization plane can be made high.

  In this way, even when the temperature difference between the target and the background is very small and the contrast is not obtained in the infrared image obtained by capturing these, the polarization characteristics (polarization degree) between the target and the background. ), The target contour can be extracted.

  From here, it demonstrates based on a specific example until the infrared target selection apparatus of Embodiment 1 extracts a target by utilizing the difference of the polarization characteristic (polarization degree) of the incident infrared rays.

  FIG. 3 is an example of an image diagram (original image) when the imaging object to be infrared-captured by the polarization image capturing apparatus 10 of the first embodiment is viewed in the visible light region. In FIG. 3, the ship that is the selection target is floating on the sea. It is assumed that the temperature difference between the ship and the sea surface is very small. At this time, when these ships and the sea surface are imaged by a normal infrared imaging device, the ship cannot be accurately extracted from the infrared image because the contrast between the ship and the sea surface cannot be obtained.

In FIG. 4A, infrared rays incident from the ship, which is the object to be imaged, and the sea surface are separated by the polarizer 51 of the polarization image capturing apparatus 10, and then imaged, added by the addition signal processing unit 53, and output. FIG. 6 is a diagram showing a luminance digital image signal 107 obtained by digitizing the luminance analog image data 106 obtained by the AD conversion circuit 12.
As shown in FIG. 4A, the luminance digital image signal 107 is composed of a signal amount for each of a plurality of pixels corresponding to each light receiving element of the infrared image sensor. The numerical value given to each pixel of the luminance digital image signal 107 in FIG. 4A is the same in the P-wave polarized image signal 100P and the S-wave polarized image signal 100S exemplified in FIGS. 5A and 5B. This is a value obtained by adding the numerical values of the P-wave polarized image signal 100P and the S-wave polarized image signal 100S at the coordinate position. As described above, the values of the pixels of the P-wave polarization image signal 100P and the S-wave polarization image signal 100S in FIGS. Is different.
FIG. 4B is an image image diagram in which the numerical value of each pixel of the luminance digital image signal 107 shown in FIG.
In this way, when the temperature difference between the ship and the sea surface is very small, the image obtained from the luminance digital image signal 107 cannot obtain the contrast between the target and the background, so the ship (target) is extracted from the image. Is difficult.

On the other hand, FIG. 6A shows the infrared rays incident from the ship, which is the imaging object, and the sea surface, after being separated by the polarizer 51 of the polarization imaging device 10, and the difference signal processing unit 52 performs the difference processing. 5 is a diagram showing a polarization difference digital image signal 102 obtained by digitizing the polarization difference analog image data 101 outputted in this manner by the AD conversion circuit 11. FIG.
The numerical value of each pixel of the polarization difference digital image signal 102 in FIG. 6A is a difference result obtained by taking a difference in signal amount between corresponding pixels in the P wave polarization image signal 100P and the S wave polarization image signal 100S. It is the indicated value. FIG. 6B is an image image diagram in which the numerical value of each pixel of the polarization difference digital image signal 102 shown in FIG.

  The polarization plane of the reflected infrared rays from the background is unknown in advance, so that the contrast between the target and the background can be obtained while rotating the polarizer 51 by the rotating unit 56, that is, the polarization difference digital image signal at the location corresponding to the background. You may adjust the rotation angle of the polarizer 51 so that the pixel value of 102 may become large. Alternatively, a device that detects the polarization plane of reflected infrared light from the background is provided, and the polarizer 51 is rotated by the rotating unit 56 so that either the background polarization plane detected by the device matches the polarization plane orthogonal to the polarizer 51 or not. You may make it rotate.

Next, the binarization circuit 13 binarizes the value of each pixel of the polarization difference digital image signal 102. In binarization, the binarization circuit 13 accesses the threshold value database 17 and selects the threshold value 109 from the threshold value database 17 based on the target to be extracted and the background type. Then, by comparing with the selected threshold 109, the value of each pixel of the polarization difference digital image signal 102 is binarized.
Note that the type of the target and its background can be designated in advance by inputting a number corresponding to the target to be selected and the type of the background using the input unit 21 such as a keyboard.

FIG. 7 is a diagram showing an example of the threshold database 17.
The threshold value database 17 is a value set by a prior survey, and the background polarization difference digital image signal in the binarized image corresponding to the type of combination of the target to be extracted and its background. A threshold 109 that can be separated from the target polarization difference digital image signal is stored in advance. For example, when the target is a ship and the background is the sea, a threshold value of a polarization difference digital image signal (threshold value = 0.6 in the example of FIG. 7) that can be separated is stored in advance.

  As described above, the binarization circuit 13 selects the threshold 109 corresponding to the type of the target and the background selected from the threshold database 17, and binarizes the polarization difference digital image signal 102 by comparison with the threshold 109. Turn into. FIG. 8 shows a binarized digital image signal 103 obtained by binarizing the polarization difference digital image signal 102 of FIG. 5A when the target is a ship and the background is the sea (threshold = 0.6 from FIG. 7). An example of Here, the binarization circuit 13 binarizes the polarization difference digital image signal 102 with a value less than the threshold 109 being 1 and a threshold 109 or more being 0.

  Next, the region feature quantity measurement unit 14 inputs the binarized digital image 103 and the luminance digital image signal 107.

  The area feature quantity measurement unit 14 performs multiplication processing on the information (numerical values) for each pixel of the input binary digital image 103 and the luminance digital image signal 107 according to Equation 2.

Here, α (i, j) is the pixel value at the coordinate position (i, j) in the two-dimensional target candidate extraction image 104, and β (i, j) is the binarized binary digital image 103. The pixel value at the coordinate position (i, j), γ (i, j) indicates the luminance value of the pixel at the coordinate position (i, j) of the luminance digital image signal 107.
According to Equation 2, α (i, j) = γ (i, j) when β (i, j) = 1, and α (i, j) = when β (i, j) = 0. 0.

FIG. 9A is a diagram showing an example of the target candidate extraction image 104 after processing according to Equation 2.
FIG. 9B is a diagram in which the numerical values of each pixel of the target candidate extraction image 104 are represented by shading, and the target candidates to be selected are extracted as shapes. The value of each pixel in the extracted target candidate area indicates the luminance information of infrared rays incident from the target candidate.

As described above, the region feature quantity measurement unit 14 erases the luminance digital signal of the background region by the multiplication processing of Expression 2, and obtains the target candidate extraction image 104 that leaves the luminance digital image signal of only the target and other target regions. Can be obtained.
Since the region feature quantity measurement unit 14 extracts the target candidates in this way, it corresponds to the target candidate extraction unit.

  As described above, according to the infrared target selection device of the first embodiment, even if the temperature difference between the target and the background is very small, the target candidate to be selected from the infrared captured image is selected. Can be extracted.

  Next, the target determination processing unit 15 collates the image with the selected target candidate and target information 108 including target shape and luminance information stored in advance in the target image database 20.

  FIG. 10 is a diagram showing an example of the target information 108 stored in the target image database 20. The target image database 20 stores information on the shape (the shape surrounded by the black line frame in FIG. 10) and the luminance (the numerical value of each pixel in the thick black line frame) for each target. In the example of FIG. 10, information on the ship 1 and the buildings 1 to 3 is stored.

  The target determination processing unit 15 compares the target candidate extraction image 104 with the target information 108 (here, the ship 1 in FIG. 10), and performs predetermined image matching on the shape and the brightness, thereby obtaining the target candidate extraction image 104. A target is selected from among them (see FIG. 11). The image matching between the target candidate extracted image 104 and the target information 108 may be, for example, pattern matching based on a difference sum value or image matching focusing on a characteristic shape. Further, collation based on the difference between the maximum luminance and the minimum luminance may be performed, and for example, image collation may be performed based on the position of the high luminance point and the number thereof.

  When the target determination processing unit 15 selects a target, the target determination processing unit 15 generates and outputs a guidance signal 105 based on the selected target image position. The flying body steers the aircraft by the guidance signal 105 and guides it on the line of sight to the target, and guides the flying body to the target.

  As described above, the infrared target selection device according to the present invention pays attention to the fact that the polarization degree of the infrared light emitted from the substance differs depending on the surface characteristics, and the polarization difference analog image data 101 corresponding to the polarization degree and the luminance analog image. A target candidate is extracted using the data 106 and the threshold 109 in the threshold database 17, and the target information is compared with the extracted target candidate and the target information 108 in the target image database 20, and the target is determined. It was made to choose.

  As a result, even if the temperature difference between the target and the background is very small, and the conventional apparatus cannot obtain the contrast between the target and the background, it is difficult to extract the target. be able to.

  The data formats of the threshold database in FIG. 7 and the target image database in FIG. 10 are examples, and other formats may be used as long as the purpose of each database is achieved. It doesn't matter. An example of the number of pixels of the image and the numerical values given here is also shown.

It is a block diagram which shows the structure of the infrared target selection apparatus by Embodiment 1 of this invention. It is a figure explaining the outline of a structure and operation | movement of the polarized image imaging device by Embodiment 1 of this invention. It is the figure which showed an example of the imaging target which the polarized image imaging device 10 by Embodiment 1 of this invention images. It is the figure which showed an example of the luminance digital image signal 107 by Embodiment 1 of this invention. It is a figure showing an example of P wave polarization image signal 100P and S wave polarization image signal 100S by Embodiment 1 of the present invention. It is the figure which showed an example of the polarization difference digital image signal 102 by Embodiment 1 of this invention. It is a figure which shows an example of the threshold value database by Embodiment 1 of this invention. It is the figure which showed an example of the binarized digital image signal 103 by Embodiment 1 of this invention. It is the figure which showed an example of the target candidate extraction image 104 by Embodiment 1 of this invention. It is a figure which shows an example of the target image database by Embodiment 1 of this invention. It is the figure which showed an example of the result as which the target determination process part 15 by Embodiment 1 of this invention extracted the target.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 Infrared target selection apparatus, 10 Polarized image imaging device, 11, 12 AD conversion circuit, 13 Binarization circuit, 14 Area | region feature-value measurement part, 15 Target determination process part, 16 Target shape / luminance image database, 17 Threshold database, 20 Target image database, 52 Difference signal processing unit, 53 Addition signal processing unit, 54 First infrared image sensor, 55 Second infrared image sensor, 56 Rotating unit, 100 Infrared input light, 90P P wave infrared, 90S S wave Infrared, 100 Infrared input light, 100P P wave polarization image signal, 100S S wave polarization image signal, 101 Polarization difference analog image data, 102 Polarization difference digital image signal, 103 Binary digital image signal, 104 Target candidate extraction image, 105 Target determination result 106 luminance analog image data, 107 luminance digital Image signal, 108 target information, 109 threshold

Claims (2)

Infrared rays incident from the target and the background of the target are separated into plane polarized light orthogonal to each other, and each of the separated infrared rays is received by an image sensor including light receiving elements arranged in two-dimensional coordinates, and electric signals are received by the light receiving elements. A polarization difference image signal obtained by taking a difference between the electric signal amounts at the same coordinate position of the two-dimensional coordinates between the respective image signals converted into a quantity and outputting the two-dimensional coordinates, A polarization image capturing apparatus that outputs a luminance image signal obtained by adding the electric signal amounts at the same coordinate position; and
A threshold database that stores in advance data associated with a combination of a target to be extracted and a background thereof, and a threshold at which the target polarization difference image signal and the background polarization difference image signal in the combination are separable ; ,
Based on a combination of the target to be extracted and its background, a binarized image signal obtained by binarizing the electric signal amount of the polarization difference image signal based on a comparison with the threshold selected from the threshold database is output. A binarization circuit;
The target candidate is extracted from an image obtained by multiplying the binarized value of the binarized image signal and the electric signal amount of the luminance image signal at the same coordinate position of the two-dimensional coordinates. An infrared target selection device comprising a target candidate extraction unit. Furthermore, a target image database storing target information which is information on the target shape and brightness,
The infrared target selection device according to claim 1, further comprising: a target determination processing unit that compares the target candidate with target information in the target image database and selects a target from the target candidates.

Images