A device and method for detecting the coordinates of a burst point in space

Abstract

The application discloses a burst point space coordinate detection device and a measuring method, which comprises an image processing system; an infrared detection device is arranged in a terminal ballistic safety area, and at least one high-speed camera is arranged on both sides of the ballistic, so that the pre-burst range of the shell is in the intersection detection field area, a test marker is placed at the theoretical burst point, and the coordinates are measured; a screen target trigger device is placed in the safety area below the front pre-ballistic at a distance of 300-500 meters from the center of the high-speed camera on both sides of the ballistic, the burst point image processing system and the Beidou time system device are arranged in the safety area outside the burst point; the high-speed camera is connected with the screen target trigger device, the burst point image processing system and the Beidou time system device respectively, and the infrared detection device and the Beidou time system device are connected with the burst point image processing system. The problem that the search of the burst point image is time-consuming and inaccurate due to the limited frame frequency of the high-speed camera and the large image storage is solved, and the precision of the burst point coordinate measurement is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of image processing technology and target detection, and particularly relates to a blast point space coordinate detection device and a measurement method. BACKGROUND

[0002] In a target range test, the measurement of blast point space coordinates is one of the most important test items of a conventional target range, and is of great significance to the damage effectiveness evaluation of a weapon system. There are three common methods for blast point space coordinate measurement technology: photoelectric measurement, acoustic sensor measurement and image measurement. Photoelectric measurement mainly uses a photoelectric theodolite to measure blast point coordinates, and has the advantages of early start, wide application and high automation. The current blast point theodolite is the main near-ground blast point measurement equipment in a target range. However, there are still some defects in practical application. Since the focal length is fixed and the lens and the camera are packaged integrally and cannot be replaced, the application range is limited for different test requirements. In addition, due to the low frame frequency of the traditional blast point theodolite, the first frame of explosion image captured is large, which affects the extraction accuracy of blast point pixel coordinates. Acoustic sensor measurement is to obtain the explosion shock wave information radiated to the surrounding when explosion, and combine a sky screen and a blast point flame detector to measure the direction of the shell flight and the three-dimensional coordinates of the blast point by using the multi-sensor information fusion theory. The array is flexible, the detection range is wide, the distance is far, and it is not affected by visibility and observation field shielding, and can be used all day and all weather. However, the explosion sound wave is greatly affected by the terrain and environment of the shell drop area, the sound wave signal is easily confused and stuck, and the positioning error is large. The image measurement method uses a high-speed camera to track and shoot the measurement target, and has the advantages of convenient station arrangement and flexible replacement of multiple lenses. In recent years, it has been applied more in target range tests. Since the explosion light of the shell expands rapidly at the moment of explosion, and smoke is generated, the time of the explosion light maintenance is only a few milliseconds, so it has high speed and instantaneousness. Limited by the frame frequency of the high-speed camera, the first moment of explosion light image information cannot be captured. SUMMARY

[0003] The present application solves the problem of providing an air blast point three-dimensional coordinate detection device and measurement method, which solves the problem of blast point space coordinate measurement due to the limitation of high-speed camera frame frequency and large image storage, and improves the accuracy of blast point coordinate measurement.

[0004] The present application is realized by the following technical solutions:

[0005] The application discloses a burst point space coordinate detection device, which comprises an infrared detection device, a sky screen trigger device, a test marker, a burst point image processing system, a Beidou time system device and at least two high-speed cameras.

[0006] The test marker is at least two groups of test markers.

[0007] The high-speed camera is used for synchronously shooting sequence image information of a target cannonball in a terminal trajectory from multiple angles and at a close distance; the sky screen trigger device is used for providing a unified trigger signal to the high-speed cameras on both sides of the trajectory, so that the high-speed cameras are provided with accurate synchronous starting shooting signals when the cannonball flies through the sky screen trigger device, the image information of the high-speed cameras is accurately shot before and after the explosion of the cannonball, and the storage capacity of the image information of the high-speed cameras is reduced; the infrared detection device is used for capturing an infrared signal of a cannonball explosion burst point and transmitting the infrared signal to the burst point image processing system to determine cannonball explosion time information, so that accurate and reliable cannonball burst point images are quickly extracted; the test marker is used for providing known coordinate points for the high-speed cameras on both sides of the trajectory before the test starts, and the internal and external parameters of the high-speed cameras are calibrated and space coordinate calculation is carried out after the test; the burst point image processing system is used for acquiring cannonball image information by adopting image framing according to the cannonball explosion time information provided by the infrared detection device, and making a judgment on the cannonball image information; if the high-speed camera shoots an image of the explosion time point at the cannonball explosion time point, the burst point space coordinate is calculated by using the image; if the high-speed camera does not shoot an image of the explosion time point at the cannonball explosion time point, a cannonball image information of a frame before the explosion and a cannonball fire image information of a first frame after the explosion are extracted, and the burst point image information of the first time of the cannonball explosion is acquired by adopting an image interpolation algorithm, combining a cannonball explosion fireball inflation model and the cannonball explosion time, so that the burst point space three-dimensional coordinate of the first time of the cannonball explosion and the explosion time information are calculated.

[0008] A burst point space coordinate measurement method using a burst point space coordinate detection device, comprising the following steps:

[0009] Step 1): The burst point image processing system receives the target projectile sequence image information sent by the high-speed camera and the projectile explosion time information sent by the infrared detection device in real time, and quickly extracts the projectile image information of the burst point region according to the projectile explosion time information provided by the infrared detection device using a video frame method, and judges the projectile image information, judges whether the high-speed camera has photographed the image at the explosion time point, and performs the following two operation modes;

[0010] Mode one: if it is judged that the high-speed camera has photographed the image at the explosion time point in the projectile image information at the explosion time point of the projectile, the image at the explosion time point is the burst point image Gd'0(u e ,v e ,t0) when the projectile is initiated, and then the background reduction and morphological filtering method is used to detect and identify the burst point image information at the projectile explosion time and the corresponding time T n0 ;

[0011] Mode two: if it is judged that the high-speed camera has not photographed the image at the explosion time point in the projectile image information at the explosion time point of the projectile, the projectile explosion time information point obtained by the infrared detection device is used, and the background reduction and morphological filtering method is used to quickly extract the projectile image information before the explosion and the projectile fireball image information appearing in the first frame after the explosion and the corresponding time T n-1 , T n1 ;

[0012] Then, the frame blending interpolation algorithm in the image interpolation algorithm is used, and combined with the projectile explosion fireball inflation model, the inserted frame image pixel points are calculated to generate the intermediate frame image Gd0(u e ,v e ,t0), that is, the burst point image information at the projectile explosion time;

[0013] The generated intermediate frame image is analyzed, and there are two cases: 1) after the projectile and the burst point image frame of the projectile initiation are mixed, there is no superposition or partial superposition, then the burst point coordinates of the intermediate frame image are the intermediate points of the projectile coordinates and the burst point image center coordinates when the projectile is initiated; 2) after the projectile and the burst point image frame of the projectile initiation are mixed, the projectile target is completely overlapped with the fireball image, then the burst point coordinates of the intermediate frame image are the burst point image center coordinates when the projectile is initiated;

[0014] Step 2): the burst point image Gd'0(u e ,v e ,t0) obtained by mode one or the intermediate frame image Gd0(ue e p0 p0

[0015] Step 3) : According to the test benchmark image, the camera internal and external calibration parameters are calculated, the system calibration data are obtained, the image pixel coordinate system, the image coordinate system, the camera coordinate system and the world coordinate system are established, the conversion relationship among the coordinate systems is combined, the spatial burst point coordinate model is solved, and the obtained burst point coordinates (u p0 p0 ) are substituted into the spatial burst point coordinate model to solve the spatial coordinates of the burst point and the explosion moment.

[0016] Before step 1), the following steps are further included:

[0017] Step 4) : The infrared detection device, the sky screen target triggering device, the burst point image processing system, the Beidou time system device, two groups of test benchmarks and at least two high-speed cameras installed on both sides of the terminal trajectory are arranged in the terminal trajectory safety area; the devices are installed to form a burst point spatial coordinate detection device.

[0018] Step 5) : When the target cannonball passes through the detection area of the sky screen target triggering device, the sky screen target triggering device outputs a triggering signal, simultaneously starts the high-speed cameras on both sides of the trajectory to collect real-time continuous high-frame frequency video images of the target cannonball, buffers the collected images to obtain target cannonball sequence image information, and sends the obtained target cannonball sequence image information to the burst point image processing system; the sky screen target triggering device also starts the infrared detection device, and the infrared detection device starts to collect the cannonball explosion moment information in real time, and sends the collected cannonball explosion moment information to the burst point image processing system.

[0019] The step 4) is specifically:

[0020] ​​​​​At least one high-speed camera is erected on both sides of the trajectory at the station position of the safety area of the terminal trajectory with a tripod, and rock protection is made, so that the pre-explosion range of the shell is in the intersection detection field area of the high-speed cameras on both sides of the terminal trajectory, the infrared detection device is arranged in the safety area of the theoretical explosion point of the terminal trajectory, and rock protection is made, so that the pre-explosion point of the shell is in the detection field area of the infrared detection device, the sky screen target triggering device is arranged in the safety area below the pre-trajectory at a distance of 300-500 meters from the center of the high-speed cameras on both sides of the trajectory, at least two sets of test markers are placed at the theoretical explosion point, and the coordinates of the test markers are measured, the explosion point image processing system and the Beidou timing device are arranged in the safety area outside the explosion point, and rock protection is made; the high-speed cameras are connected with the sky screen target triggering device, the explosion point image processing system and the Beidou timing device, and the infrared detection device and the Beidou timing device are connected with the explosion point image processing system.

[0021] Step 5) is specifically: when the target shell passes through the light curtain formed by the optical lens of the sky screen target of the sky screen target triggering device detection area, the target shell blocks part of the light, so that the light flux on the photosensitive surface of the photoelectric sensor on the optical lens changes, the changed light flux is processed by the analog circuit to extract, amplify, noise filter and level conversion circuit, and finally a fixed pulse width TTL level signal is output to the high-speed cameras on both sides of the trajectory, the high-speed cameras on both sides of the trajectory are started to collect real-time continuous high-frame frequency video images of the target shell, the collected images are cached, target shell sequence image information is obtained, and the target shell sequence image information is sent to the explosion point image processing system; the sky screen target triggering device also starts the infrared detection device, and the infrared detection device starts to collect the shell explosion time information and sends the collected shell explosion time information to the explosion point image processing system.

[0022] The step 1) is specifically:

[0023] Step 1): the explosion point image processing system receives the target shell sequence image information sent by the high-speed camera and the shell explosion time information sent by the infrared detection device in real time, and according to the shell explosion time information provided by the infrared detection device, the start and end time of the target shell video information before and after the shell explosion in the target shell sequence image information is quickly extracted, the video frame method is used to derive each frame of JPEG image I(u e ,v e ,t) with time mark, then the shell image information of the explosion point area is quickly extracted, and a judgment is made on the shell image information, whether the high-speed camera has shot the image of the explosion time point, and the following two operation modes are executed.

[0024] Method 1: If it is determined that the high-speed camera captured an image at the time of the shell's explosion, then the image at the explosion time is the image of the shell's detonation point Gd'0(u) e ,v e ,t0), acquire the image of the frame before the shell enters the field of view as the background image M. bg (u e ,v e ,t0); Gd'0(u) is the image of the detonation point of a time-stamped shell. e ,v e Background subtraction algorithm is used to subtract background data from the image of the explosion point of the shell at multiple gray levels after background subtraction, and then the image of the explosion point is processed. e ,v e The image Gd'0(u) is binarized, and the threshold is determined using the maximum inter-class variance method. Pixel values ​​greater than or equal to the threshold are set to 1, and pixel values ​​less than the threshold are set to 0. Then, morphological filtering is used to refine the binarized image Gd'0(u) at the detonation point. e ,v e The ,t0) process performs an opening operation filter with erosion followed by dilation to remove small particle noise, smooth the target boundary, and accurately extract the explosion point target or shell target and flash target, obtaining the explosion point image information Gd at the moment of shell explosion. n0 (u e ,v e ,t n0 ) and corresponding time T n0 ;

[0025] Method 2: If it is determined that the high-speed camera did not capture an image at the time of the shell explosion in the shell image information, then the explosion time information obtained by the infrared detection device is used to quickly extract the target shell sequence image information from the shell image information, and the image of the frame before the shell enters the field of view is obtained as the background image M. bg (u e ,v e The target shell sequence image information with time stamps is processed by background subtraction algorithm to reduce background. Then, the multi-grayscale target shell sequence image information after background subtraction is binarized. The threshold is determined by the maximum inter-class variance method. Pixel values ​​greater than or equal to the threshold are set to 1, and pixel values ​​less than the threshold are set to 0. Then, the morphological filtering method is used to perform opening operation filtering on the binarized target shell sequence image information with erosion and dilation to remove small particle noise, smooth the target boundary, and not change its shape and area. The detonation point target or shell target and flash target are accurately extracted to obtain the shell image information Gd of the frame before detonation. n-1 (ue ,v e ,t n-1 ) and the shell flash image information when the first frame after the shell explodes. Gd n1 (u e ,v e ,t n1 ) and the corresponding time T for each frame n-1 T n1 ;

[0026] Then, the morphologically filtered image of the shell before detonation, Gd, is used. n-1 (u e ,v e ,t n-1 The image of the shell's flash in the first frame after the shell explodes (Gd) n1 (u e ,v e ,t n1 Substituting these values ​​into the frame interpolation algorithm, and combining them with the fireball expansion model of a shell explosion, assuming the weight of the frame before detonation is (L-α) / L, and the weight of the first frame's fire image after detonation is α / L, the intermediate frame image Gd0(u) is obtained by multiplying the two reference frames by their respective weights and then adding them together. e ,v e ,t0); where the algorithm is as follows:

[0027]

[0028] Where L is the distance between the two original frames, and α is the relative distance between the frame before detonation and the inserted frame;

[0029] The insertion frame time is:

[0030] T n0 =T n-1 +α

[0031] In Method 2, the generated intermediate frame images are analyzed, and two cases exist:

[0032] (1) If the shell and the flash image are superimposed or partially superimposed, the explosion point coordinates of the intermediate frame image are the midpoint between the shell coordinates and the center pixel coordinates of the flash image; (2) If the shell target completely overlaps with the flash image, the explosion point coordinates of the intermediate frame image are the center pixel coordinates of the flash image.

[0033] Step 2) specifically involves:

[0034] The image of the explosion point when the shell detonates, obtained by method one, is Gd'0(u e ,v e The intermediate frame image Gd0(u) obtained by method 2 is either t0) or t0(u). e ,ve The coordinates of the explosion point (u) are calculated using a moment-based centroid coordinate extraction algorithm. p0 ,v p0 ); Assume the mass of each pixel within the exploded region of the image is 1, meaning the mass of each pixel is equal to its pixel value, (u e ,v e Let P be the coordinates of a pixel in the image, and S be the area of ​​the pixel region. Then, the p+q moment of the target can be expressed as:

[0035]

[0036] In the formula, M represents the moments of the image at different p and q values, and f(u) e ,v e Let be the quality of a pixel; calculate its zeroth and first moments, with three cases as follows:

[0037] When p = 0 and q = 0, the zeroth moment M(0,0) can be obtained as:

[0038]

[0039] When p = 1 and q = 0, the first moment M(0,1) is:

[0040]

[0041] When p = 0 and q = 1, the value of the first moment M(1,0) is:

[0042]

[0043] The centroid of the target image can be calculated using the zeroth and first moments, using (u p0 ,v p0 Let ) represent the coordinates of the explosion point. The algorithm for solving the explosion point coordinates is as follows:

[0044]

[0045] Where M(1,0) represents the sum of the x-coordinates of all pixels in the bullet hole, M(0,1) represents the sum of the column coordinates of all pixels in the bullet hole, and M(0,0) represents the number of pixels contained in the bullet hole.

[0046] Step 3) specifically involves: solving the spatial explosion point coordinate model and substituting the explosion point coordinates into it to solve for the three-dimensional spatial coordinates of the explosion point.

[0047] Assume the three-dimensional coordinates of the spatial explosion point P in the world coordinate system are (X... P ,Y P Z P The coordinates of the image after imaging are (u P ,v P) ; the high-speed camera linear model can be expressed as:

[0048]

[0049] The camera distortion model is as follows:

[0050]

[0051] wherein, Considering the second-order radial distortion, the distortion coefficients are a1 and a2; λ is a scale factor, (u w ,v w ) is the distortion-free image coordinate, (R, T) is the external parameter of the camera, R and T are respectively the rotation matrix and the translation vector of the world coordinate system to the camera coordinate system, and A is the internal parameter matrix of the camera, which can be expressed as:

[0052]

[0053] In the formula, (u0, v0) is the principal point coordinate of the image coordinate system, f x , f y are the scale factors of the u-axis and the v-axis respectively, and α is the non-perpendicular factor of the u-axis and the v-axis; the high-speed camera calibration needs to refer to the five parameters f x , f y , α, u0 and v0 of the camera internal parameter matrix, and the distortion coefficients a1 and a2;

[0054] Before the test, two standard test poles are rotated in multiple directions within the range of the shell pre-explosion point, and a plurality of standard test pole images in different directions are shot by using the high-speed binocular camera; according to the imaging relationship of the feature points on the standard test pole on the binocular camera composed of the two high-speed cameras, the five parameters of the camera calibration and the camera distortion coefficients are calculated;

[0055] By establishing the image pixel coordinate system, the image coordinate system, the camera coordinate system and the world coordinate system, and combining the conversion relationship between the coordinate systems, the conversion relationship from the pixel coordinate system to the world coordinate system is obtained as follows:

[0056]

[0057] wherein, the value of Z c is 1, s is the coordinate axis inclination parameter, which is 0 in the ideal case, A is the camera internal parameter matrix, the orthogonal rotation matrix R is the cosine combination of the direction of the camera coordinate system relative to the coordinate axis of the world coordinate system, and the translation matrix T = [t1 t2 t3] T is the coordinate of the camera coordinate system origin in the world coordinate system; further calculation of the above formula can obtain the conversion relationship between the system pixel coordinate and the world coordinate system as follows:

[0058]

[0059] Let The above formula can be simplified to the following space burst point coordinate model:

[0060] [X Y Z 1] T =C -1 [u v 1] T

[0061] In the formula, (X, Y, Z) is the solved space burst point coordinate, and the unit is meter; (u, v) is the pixel coordinate, and the unit is pixel;

[0062] Wherein, the burst point coordinate (u p0 ,v p0 ) obtained in step 2) is substituted into the above formula space burst point coordinate model to obtain the space three-dimensional coordinates (X, Y, Z) of the burst point at the explosion moment.

[0063] Compared with the prior art, the present application has the following beneficial technical effects:

[0064] The burst point space coordinate detection device and the measurement method provided by the present application can quickly extract the shell image information of the burst point region from the shell sequence image information captured by the high-speed camera based on the shell explosion moment information provided by the infrared detection device, and can directly obtain the burst point image information after judgment, and then calculate the space three-dimensional coordinates of the air burst point, thereby improving the work efficiency and reducing the labor intensity, or can quickly extract the shell image information before explosion and the shell fire image information after explosion from the shell image information, thereby quickly finding the burst point coordinates, improving the work efficiency and reducing the labor intensity. Meanwhile, the binocular high-speed camera detection mechanism is used to fully utilize the relationship between the sequence images of the camera shooting burst point formation process, the image frame insertion method is used to generate more accurate first frame fire image of the burst point explosion moment, the continuous frame image is formed, the binocular vision detection principle is combined, the space three-dimensional coordinates of the air burst point are calculated, and the problem that the camera frame frequency is limited and the fire image information of the initial explosion moment of the shell cannot be captured is solved, thereby improving the calculation accuracy of the air burst point coordinates.

[0065] The application provides a blast point space coordinate detection device and a measurement method, and solves the technical problem that in the prior art, the explosion light diffuses very fast, is limited by the frame frequency of a high-speed camera, and cannot capture image information of the initial moment of explosion of the explosion fireball, and the image of the explosion fireball in the first frame after explosion is often a large fireball, and even has changed into an irregular fireball, which leads to a large deviation in the calculation of the instantaneous coordinates of the explosion fireball. BRIEF DESCRIPTION OF DRAWINGS

[0066] Figure 1 is a detection system station layout diagram implemented by the application;

[0067] Figure 2 is a blast point image interpolation principle of the application;

[0068] Figure 3 is an algorithm flowchart of the application;

[0069] Figure 4 is an image time relationship diagram of the interpolation algorithm of the application;

[0070] Figure 5 is a conversion relationship diagram of the camera coordinate system and the world coordinate system. DETAILED DESCRIPTION

[0071] The application will be further described in detail below in combination with specific embodiments, and the description is an explanation instead of a limitation of the application.

[0072] Embodiment 1

[0073] Reference Figure 1The application discloses a burst point space coordinate detection device, which comprises an infrared detection device, a sky screen target triggering device, a burst point image processing system, a Beidou time system device, two groups of test markers and at least two high-speed cameras.

[0074] The high-speed camera is used for synchronously shooting the sequence image information of the target cannonball in the terminal trajectory from multiple angles and at a close distance; the sky screen target triggering device is used for providing a unified triggering signal to the high-speed cameras on both sides of the trajectory, so as to ensure that the high-speed cameras are provided with accurate synchronous starting shooting signals when the cannonball flies through the sky screen target triggering device, and to make the high-speed cameras accurately shoot the image information in the period before and after the explosion of the cannonball, thereby reducing the storage capacity of the image information of the high-speed camera; the infrared detection device is used for capturing the infrared signal of the explosion point of the cannonball and transmitting the infrared signal to the burst point image processing system to determine the explosion time information of the cannonball, so as to quickly extract the accurate and reliable cannonball burst point image; the test marker is used for providing the known coordinate points for the high-speed cameras on both sides of the trajectory before the test starts, and calibrating the internal and external parameters of the high-speed camera and solving the space coordinates after the test; the burst point image processing system is used for acquiring the cannonball image information by using image framing according to the explosion time information of the cannonball provided by the infrared detection device, and judging the cannonball image information; if the high-speed camera shoots the image of the explosion time point at the explosion time point of the cannonball, the frame image is used to calculate the burst point space coordinates; if the high-speed camera does not shoot the image of the explosion time point at the explosion time point of the cannonball, the cannonball image information of the frame before the explosion and the cannonball fire image information of the first frame after the explosion are extracted, and the image interpolation algorithm is used to combine the cannonball explosion fireball inflation model and the explosion time of the cannonball to obtain the burst point image information at the first time of the cannonball initiation, so as to calculate the burst point space three-dimensional coordinates at the first time of the cannonball explosion and the explosion time information.

[0075] Referring to Figure 1 to toFigure 5 A burst point space coordinate measurement method using the burst point space coordinate detection device, comprising the following steps:

[0076] Step 1): The infrared detection device, the sky screen target triggering device, the burst point image processing system, the Beidou time system device, two groups of test markers and at least two high-speed cameras installed on both sides of the terminal trajectory are arranged in the safety area of the terminal trajectory; each device is installed to form a burst point space coordinate detection device;

[0077] The step 1) is specifically:

[0078] At least one high-speed camera is erected on both sides of the trajectory at the station position in the safety area of the terminal trajectory with a tripod, and rock protection is performed, so that the pre-burst range of the shell is in the intersection detection field area of the high-speed cameras on both sides of the trajectory, the infrared detection device is arranged in the safety area of the theoretical burst point of the trajectory, and rock protection is performed, so that the pre-burst point of the shell is in the detection field area of the infrared detection device, the sky screen target triggering device is arranged in the safety area below the pre-trajectory at a distance of 300-500 meters from the center of the high-speed cameras on both sides of the trajectory, at least two groups of test markers are placed at the theoretical burst point, and the coordinates of the test markers are measured, the burst point image processing system and the Beidou time system device are arranged in the safety area outside the burst point, and rock protection is performed; the high-speed cameras are connected with the sky screen target triggering device, the burst point image processing system and the Beidou time system device respectively, and the infrared detection device, the Beidou time system device and the burst point image processing system are connected.

[0079] Step 2): When the target shell passes through the detection area of the sky screen target triggering device, the sky screen target triggering device outputs a triggering signal, simultaneously starts the high-speed cameras on both sides of the trajectory to collect real-time continuous high-frame frequency video images of the target shell, buffers the collected images to obtain target shell sequence image information, and sends the obtained target shell sequence image information to the burst point image processing system; the sky screen target triggering device also simultaneously starts the infrared detection device, and the infrared detection device starts to collect the shell explosion time information in real time and sends the collected shell explosion time information to the burst point image processing system.

[0080] Step 2) is specifically: when the target projectile passes through the light curtain formed by the optical lens of the screen target trigger device in the detection area of the screen target, the target projectile blocks part of the light, causing the light flux on the photosensitive surface of the photoelectric sensor reaching the optical lens of the screen target to change, the changed light flux is processed by the analog circuit to extract, amplify, noise filter, and level conversion circuit, and finally outputs a fixed pulse width TTL level signal to the high-speed cameras on both sides of the trajectory, starts the high-speed cameras on both sides of the trajectory to collect real-time continuous high-frame frequency video images of the target projectile, caches the collected images, obtains the target projectile sequence image information, and sends the obtained target projectile sequence image information to the burst point image processing system; the screen target trigger device also starts the infrared detection device, and the infrared detection device starts to collect the projectile explosion time information in real time and sends the collected projectile explosion time information to the burst point image processing system.

[0081] Step 3): the burst point image processing system receives the target projectile sequence image information sent by the high-speed camera and the projectile explosion time information sent by the infrared detection device in real time, and quickly extracts the projectile image information of the burst point area according to the projectile explosion time information provided by the infrared detection device using a video frame extraction method, and judges the projectile image information, judges whether the high-speed camera has photographed the image of the explosion time point, and performs the following two operation modes;

[0082] Mode one: if it is judged that the high-speed camera has photographed the image of the explosion time point in the projectile image information at the time point of the explosion of the projectile, the image of the explosion time point is the burst point image Gd'0(u e ,v e ,t0) of the projectile at the time of initiation, and then the background reduction and morphological filtering method is used to detect and identify the burst point image information at the time of the explosion of the projectile and the corresponding time T n0 ;

[0083] Mode two: if it is judged that the high-speed camera has not photographed the image of the explosion time point in the projectile image information at the time point of the explosion of the projectile, the projectile explosion time information point obtained by the infrared detection device is used, and the background reduction and morphological filtering method is used to quickly extract the projectile image information before the explosion of the first frame and the projectile fire image information after the explosion of the first frame of fire light in the projectile image information and the corresponding time T n-1 ,T n1 ;

[0084] Then, the frame blending interpolation algorithm in the image interpolation algorithm is used, and combined with the explosion fireball expansion model of the projectile, the pixel points of the inserted frame image are calculated, and the intermediate frame image Gd0(u e ,v e ,t0) is generated, that is, the burst point image information at the time of the explosion of the projectile;

[0085] The generated intermediate frame image is analyzed, and there are two cases: 1) the bomb and the bomb explosion point image frame are mixed, and there is no superposition or partial superposition, so that the explosion point coordinates of the intermediate frame image are the middle points of the bomb coordinates and the bomb explosion point image center coordinates; 2) the bomb and the bomb explosion point image frame are mixed, and the bomb target is completely overlapped with the fire image, so that the explosion point coordinates of the intermediate frame image are the center coordinates of the bomb explosion point image.

[0086] The step 3) is specifically:

[0087] Step 3): The explosion point image processing system receives the target bomb sequence image information sent by the high-speed camera and the bomb explosion time information sent by the infrared detection device in real time, and according to the bomb explosion time information provided by the infrared detection device, the start and end time of the target bomb video information before and after the bomb explosion in the target bomb sequence image information is quickly extracted, the video frame method is used to derive each frame of JPEG image I(u e ,v e ,t) with time mark, then the bomb image information of the explosion point area is quickly extracted, and the bomb image information is judged to determine whether the high-speed camera has captured the image at the explosion time point, and the following two operation modes are executed.

[0088] Mode one: if it is judged that the high-speed camera has captured the image at the explosion time point in the bomb image information at the explosion time point, the image at the explosion time point is the bomb explosion point image Gd'0(u e ,v e ,t0), and a frame of image before the bomb enters the field of view is obtained as a background image M bg (u e ,v e ,t0); the background difference algorithm is used to perform background subtraction processing on the bomb explosion point image Gd'0(u e ,v e ,t0) with time mark, and then the bomb explosion point image Gd'0(u e ,v e ,t0) with multiple gray scales after the background subtraction processing is subjected to binaryzation processing, the threshold value is determined by using the maximum inter-class variance method, the pixel value greater than or equal to the threshold value is set to 1, and the pixel value less than the threshold value is set to 0; then the morphological filtering method is used to perform the opening operation filtering processing of first corrosion and then inflation on the bomb explosion point image Gd'0(u e ,v e ,t0) after the binaryzation processing, small particle noise is removed, the target boundary is smoothed, and the shape and area are not changed, the explosion point target or the bomb target and the fire target are accurately extracted, and the explosion point image information Gd n0 (ue ,v e ,t n0 ) and corresponding time T n0 ;

[0089] Method two: if it is judged that the high-speed camera does not capture the image of the explosion time point in the shell image information at the time point of the explosion of the shell, the time point information of the explosion of the shell obtained by the infrared detection device is used to quickly extract the target shell sequence image information in the shell image information, and the image before the shell enters the field of view is obtained as the background image M bg (u e ,v e ,t0); the background difference algorithm is used for background subtraction processing on the target shell sequence image information with time mark, and then the binaryzation processing is performed on the multi-gray scale target shell sequence image information after the background subtraction processing, the threshold value is determined by using the maximum inter-class variance method, the pixel value greater than or equal to the threshold value is set to 1, and the pixel value less than the threshold value is set to 0; then the morphological filtering method is used for the open operation filtering processing of the erosion and then the inflation on the binaryzation processed target shell sequence image information, the small particle noise is removed, the target boundary is smoothed, and the shape and area are not changed, the blast point target or the shell target and the fire light target are accurately extracted, and the shell image information Gd n-1 (u e ,v e ,t n-1 ) before the explosion of the previous frame and the shell fire light image information Gd n1 (u e ,v e ,t n1 ) of the first frame after the explosion of the shell appear fire light and the corresponding time T n-1 , T n1 ;

[0090] Then, the morphological filtering processed shell image Gd n-1 (u e ,v e ,t n-1 ) before the explosion of the previous frame and the shell fire light image Gd n1 (u e ,v e ,t n1 ) of the first frame after the explosion of the shell appear fire light are brought into the interpolation algorithm, and the shell explosion fireball inflation model is combined, it is assumed that the weight of the previous frame before the explosion is (L-α) / L, and the weight of the first frame fire light image after the explosion is α / L, the two reference frames are added after being multiplied by the respective weights, and the intermediate frame image Gd0(u e ,v e ,t0) can be obtained; wherein, the algorithm is as follows:

[0091]

[0092] Where L is the distance between the two original frames, and α is the relative distance between the frame before detonation and the inserted frame;

[0093] The insertion frame time is:

[0094] T n0 =T n-1 +α

[0095] In Method 2, the generated intermediate frame images are analyzed, and two cases exist:

[0096] (1) If the shell and the flash image are superimposed or partially superimposed, the explosion point coordinates of the intermediate frame image are the midpoint between the shell coordinates and the center pixel coordinates of the flash image; (2) If the shell target completely overlaps with the flash image, the explosion point coordinates of the intermediate frame image are the center pixel coordinates of the flash image.

[0097] Step 4): Obtain the explosion point image Gd'0(u) obtained from Method 1 at the time of shell detonation. e ,v e The intermediate frame image Gd0(u) obtained by method 2 is either t0) or t0(u). e ,v e The coordinates of the explosion point (u) are calculated using a moment-based centroid coordinate extraction algorithm. p0 ,v p0 ).

[0098] Step 4) specifically involves:

[0099] The image of the explosion point when the shell detonates, obtained by method one, is Gd'0(u e ,v e The intermediate frame image Gd0(u) obtained by method 2 is either t0) or t0(u). e ,v e The coordinates of the explosion point (u) are calculated using a moment-based centroid coordinate extraction algorithm. p0 ,v p0 ); Assume the mass of each pixel within the exploded region of the image is 1, meaning the mass of each pixel is equal to its pixel value, (u e ,v e Let P be the coordinates of a pixel in the image, and S be the area of ​​the pixel region. Then, the p+q moment of the target can be expressed as:

[0100]

[0101] In the formula, M represents the moments of the image at different p and q values, and f(u) e ,v e Let be the quality of a pixel; calculate its zeroth and first moments, with three cases as follows:

[0102] When p = 0, q = 0, the zero-order moment M(0, 0) is obtained as:

[0103]

[0104] When p = 1, q = 0, the first-order moment M(0, 1) is:

[0105]

[0106] When p = 0, q = 1, the value of the first-order moment M(1, 0) is:

[0107]

[0108] The center of gravity of the target image can be calculated using the zero-order moment and the first-order moment, and the coordinates of the bomb point are represented by (u p0 ,v p0 ), so the solving algorithm of the bomb point coordinates is:

[0109]

[0110] Wherein, M(1, 0) represents the sum of the horizontal coordinates of all pixels of the bullet hole, M(0, 1) represents the sum of the column coordinates of all pixels of the bullet hole, and M(0, 0) represents the number of pixels contained in the bullet hole.

[0111] Step 5) According to the test benchmark image, the camera internal and external calibration parameters are solved, the system calibration data is obtained, the image pixel coordinate system, the image coordinate system, the camera coordinate system and the world coordinate system are established, the conversion relationship between the coordinate systems is combined, the spatial bomb point coordinate model is solved, and the obtained bomb point coordinates (u p0 ,v p0 ) are substituted into the spatial bomb point coordinate model to solve the spatial coordinates of the bomb point and the explosion time.

[0112] Step 5) is specifically: solving the spatial bomb point coordinate model and bringing the bomb point coordinates into it to solve the spatial three-dimensional coordinates of the bomb point:

[0113] Suppose the three-dimensional coordinates of the spatial bomb point P in the world coordinate system are (X P ,Y P ,Z P ), and the image coordinates after imaging are (u P ,v P ); the linear model of the high-speed camera can be expressed as:

[0114]

[0115] The camera distortion model is as follows:

[0116]

[0117] wherein, Considering the second order radial distortion, the distortion coefficients are a1, a2; λ is a scale factor, (u w ,v w ) is the undistorted image coordinate, (R, T) is the external parameter of the camera, R and T are respectively the rotation matrix and the translation vector of the world coordinate system to the camera coordinate system, and A is the internal parameter matrix of the camera, which can be expressed as:

[0118]

[0119] In the formula, (u0, v0) is the principal point coordinate of the image coordinate system, f x , f y are the scale factors of the u-axis and the v-axis respectively, and α is the non-perpendicular factor of the u-axis and the v-axis; the high-speed camera calibration needs to refer to the five parameters f x , f y , α, u0 and v0 of the internal parameter matrix of the camera, and the distortion coefficients a1 and a2;

[0120] Before the test, two standard test poles are rotated in multiple directions in the range of the projectile pre-explosion point, and a plurality of standard test pole images in different directions are shot by the high-speed binocular camera; the five parameters of the camera calibration and the camera distortion coefficients are calculated according to the imaging relationship of the feature points on the standard test pole on the binocular camera composed of the two high-speed cameras;

[0121] By establishing the image pixel coordinate system, the image coordinate system, the camera coordinate system and the world coordinate system, and combining the conversion relationship between the coordinate systems, the conversion relationship from the pixel coordinate system to the world coordinate system is obtained as follows:

[0122]

[0123] wherein, the value of Z c is 1, s is the coordinate axis tilt parameter, which is 0 in the ideal case, A is the internal parameter matrix of the camera, the orthogonal rotation matrix R is the cosine combination of the direction of the camera coordinate system relative to the coordinate axis of the world coordinate system, and the translation matrix T = [t1 t2 t3] T is the coordinate of the camera coordinate system origin in the world coordinate system; the conversion relationship between the system pixel coordinate and the world coordinate system can be obtained by further solving the above formula as follows:

[0124]

[0125] Let The above formula can be simplified to the following space explosion point coordinate model:

[0126] [X Y Z 1] T =C -1[u v 1] T

[0127] In the formula, (X, Y, Z) is the solved space burst point coordinates, and the unit is meter; (u, v) is the pixel coordinates, and the unit is pixel;

[0128] Wherein, the burst point coordinates (u p0 ,v p0 ) obtained in step 2) are substituted into the above formula space burst point coordinate model to obtain the space three-dimensional coordinates (X, Y, Z) of the burst point at the explosion moment.

[0129] The burst point space coordinate detection device and measurement method provided by the application can quickly extract the shell image information of the burst point region from the shell sequence image information captured by the high-speed camera based on the information provided by the infrared detection device at the shell explosion moment, and can directly obtain the burst point image information after judgment, and then calculate the space three-dimensional coordinates of the air burst point, thereby improving the work efficiency and reducing the labor intensity, or can quickly extract the shell image information before explosion and the shell fire image information after explosion from the shell image information, thereby quickly finding the burst point coordinates, improving the work efficiency and reducing the labor intensity. Meanwhile, the double-eye high-speed camera detection mechanism is used to fully utilize the relationship between the sequence images of the camera shooting burst point formation process, the image frame insertion method is used to generate more accurate first frame fire image at the burst point explosion moment, the continuous frame image is formed, the double-eye vision detection principle is combined, the space three-dimensional coordinates of the air burst point are calculated, and the problem that the camera frame frequency is limited and the fire image information at the initial burst moment of the explosion fireball cannot be captured is solved, thereby improving the calculation accuracy of the air burst point coordinates.

[0130] The burst point space coordinate detection device and measurement method provided by the application solves the problems in the prior art that the explosion fireball spreads very fast, the frame frequency of the high-speed camera is limited, the image information at the initial burst moment of the explosion fireball cannot be captured, and the image of the explosion fireball in the first frame after burst is a very large fireball, even changes into an irregular fireball, thereby causing a large deviation in the calculation of the burst moment coordinates of the fireball. The application can generate coherent intermediate images, predict the burst point image information of the explosive, calculate the space three-dimensional coordinates of the burst point by using the center of gravity extraction algorithm based on the matrix, solve the technical problems that the space coordinate measurement of the burst point is affected by the frame frequency limitation of the high-speed camera, and the problems that the image storage amount is large, the search of the burst point image is time-consuming and inaccurate, and the measurement precision of the burst point coordinates is improved.

[0131] The above shows and describes the basic principles and main features of the present application and the advantages of the present application. Those skilled in the art should understand that the present application is not limited to the above embodiments, and the above embodiments and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A device for detecting the coordinates of a point in space, characterized in that it comprises: The application relates to a test device for a terminal trajectory of a cannonball, which comprises an infrared detection device, a sky screen target triggering device, a burst point image processing system, a Beidou time system device, a test marker and at least two high-speed cameras; at least one high-speed camera is arranged on each side of the terminal trajectory at a station position in a safe area by a tripod, and rock protection is arranged on the high-speed cameras, so that a pre-explosion range of the cannonball is in an intersection detection field area of the high-speed cameras on both sides of the terminal trajectory; the infrared detection device is arranged in a safe area of a theoretical burst point of the terminal trajectory by a tripod, and rock protection is arranged on the infrared detection device, so that a pre-explosion point of the cannonball is in a detection field area of the infrared detection device; the sky screen target triggering device is arranged in a safe area below a pre-trajectory 300-500 m in front of a central position of the high-speed cameras on both sides of the terminal trajectory; the test marker is placed at the theoretical burst point, and the coordinates of the test marker are measured; the burst point image processing system and the Beidou time system device are arranged in a safe area outside the burst point, and rock protection is arranged on the burst point image processing system and the Beidou time system device; the high-speed cameras are connected with the sky screen target triggering device, the burst point image processing system and the Beidou time system device; the infrared detection device and the Beidou time system device are connected with the burst point image processing system; the high-speed cameras are used for synchronously shooting sequence image information of a target cannonball in the terminal trajectory from multiple angles and at a close distance; the sky screen target triggering device is used for providing a unified triggering signal to the high-speed cameras on both sides of the terminal trajectory, so as to ensure that the high-speed cameras are provided with accurate synchronous starting shooting signals when the cannonball flies through the sky screen target triggering device, and the high-speed cameras accurately shoot image information in a period before and after the explosion of the cannonball, so as to reduce the storage capacity of the high-speed cameras; The infrared detection device is used for capturing an infrared signal of a cannonball explosion burst point and transmitting the infrared signal to the burst point image processing system to determine cannonball explosion time information, so as to quickly extract accurate and reliable cannonball burst point images; the test marker is used for providing known coordinate points for the high-speed cameras on both sides of the terminal trajectory before the test starts, and the test marker is used for calibrating internal and external parameters of the high-speed cameras and solving space coordinates after the test; the burst point image processing system is used for acquiring cannonball image information by adopting image framing according to the cannonball explosion time information provided by the infrared detection device, and judging the cannonball image information; If the high-speed camera shoots an image of the explosion time point at the cannonball explosion time point, a frame image is adopted to calculate a burst point space coordinate; if the high-speed camera does not shoot an image of the explosion time point at the cannonball explosion time point, a cannonball image information of a frame before the explosion and a cannonball fire image information of a first frame after the explosion are extracted, and an image interpolation algorithm is adopted to combine a cannonball explosion fireball inflation model and the cannonball explosion time to obtain a burst point image information of a first time when the cannonball is initiated, so as to calculate a burst point space three-dimensional coordinate of the first time when the cannonball is initiated and the explosion time information.

2. The burst space coordinate probing apparatus of claim 1, wherein The test marker is at least two groups of test markers.

3. A method for measuring the coordinates of a burst point in space using the apparatus for detecting the coordinates of a burst point in space according to any one of claims 1 to 2, characterized by, The test device comprises the following steps: Step 1): The burst point image processing system receives the target projectile sequence image information sent by the high-speed camera and the projectile explosion time information sent by the infrared detection device in real time, and quickly extracts the projectile image information of the burst point region according to the projectile explosion time information provided by the infrared detection device using the video frame method, and makes a judgment on the projectile image information, judges whether the high-speed camera has photographed the image at the explosion time point, and performs the following two operation modes; The first mode: if it is judged that the high-speed camera has taken the image at the explosion time point in the shell image information, the image at the explosion time point is the explosion point image Gd'0(u e ,v e ,t0) at the shell explosion time, and then the background reduction and morphological filtering method is adopted to detect and identify the explosion point image information at the shell explosion time and the corresponding time T n0 ; Method two: if it is judged that the high-speed camera does not capture the image of the explosion time point in the shell image information, the time point information of the shell explosion obtained by the infrared detection device is used, and the background reduction and morphological filtering method is adopted to quickly extract the shell image information before explosion in the previous frame, the shell fire image information after explosion in the first frame and the corresponding time T of each frame n-1 , n1 ; Then, the frame blending interpolation algorithm in the image interpolation algorithm is adopted, and combined with the fireball expansion model of the shell explosion, the pixel points of the inserted frame image are calculated to generate the intermediate frame image Gd0(u e ,v e ,t0), that is, the burst point image information at the moment of the shell explosion; The generated intermediate frame image is analyzed, and there are two cases: 1) The projectile and the burst point image frame at the projectile initiation time are mixed, and there is no superposition or partial superposition, then the burst point coordinates of the intermediate frame image are the intermediate points of the projectile coordinates and the burst point image center coordinates at the projectile initiation time; 2) The projectile and the burst point image frame at the projectile initiation time are mixed, and the target projectile is completely overlapped with the fire image, then the burst point coordinates of the intermediate frame image are the burst point image center coordinates at the projectile initiation time; Step 2): calculate the coordinates of the burst point (u, v) by using the center of mass coordinate extraction algorithm based on the rectangle, with the image Gd'(u, v, t0) obtained in step 1) or the intermediate frame image Gd0(u, v, t0) obtained in step 2) of mode two. e e e e p0 p0 ​​​​​​ Step 3): The camera internal and external calibration parameters are calculated according to the test target image, and the system calibration data is obtained; By establishing image pixel coordinate system, image coordinate system, camera coordinate system and world coordinate system, combined with the conversion relationship between each coordinate system, the space burst point coordinate model is solved, and the obtained burst point coordinates (u p0 ,v p0 ) are substituted into the space burst point coordinate model to solve the space coordinates of the burst point and the explosion time.

4. The burst space coordinate measurement method of claim 3, wherein, Before step 1), the following steps are also included: Step 4): An infrared detection device, a screen target trigger device, a burst point image processing system, a Beidou timing device, two groups of test targets and at least two high-speed cameras installed on both sides of the terminal trajectory are arranged in the safety area of the terminal trajectory; each device is installed to form a burst point space coordinate detection device; Step 5): When the target projectile passes through the detection area of the screen target trigger device, the screen target trigger device outputs a trigger signal, and simultaneously starts the high-speed cameras on both sides of the trajectory to collect real-time continuous high-frame-rate video images of the target projectile, and buffers the collected images to obtain target projectile sequence image information, and sends the obtained target projectile sequence image information to the burst point image processing system; the screen target trigger device also starts the infrared detection device, and the infrared detection device starts to collect the projectile explosion time information in real time, and sends the collected projectile explosion time information to the burst point image processing system.

5. The burst space coordinate measurement method of claim 4, wherein, The step 4) is specifically: At least one high-speed camera is erected on both sides of the trajectory at the station position in the safety area of the terminal trajectory by using a tripod, and rock protection is performed, so that the projectile pre-explosion range is in the intersection detection field area of the high-speed cameras on both sides of the terminal trajectory, the infrared detection device is arranged in the safety area of the theoretical burst point of the terminal trajectory, and rock protection is performed, so that the projectile pre-explosion point is in the detection field area of the infrared detection device, the screen target trigger device is arranged in the safety area below the pre-trajectory at a distance of 300-500 meters from the center of the high-speed cameras on both sides of the trajectory, at least two groups of test targets are placed at the theoretical burst point, and the coordinates of the test targets are measured, the burst point image processing system and the Beidou timing device are arranged in the safety area outside the burst point, and rock protection is performed; the high-speed cameras are connected with the screen target trigger device, the burst point image processing system and the Beidou timing device, and the infrared detection device, the Beidou timing device and the burst point image processing system are connected; Step 5) is specifically: when the target projectile passes through the light curtain formed by the optical lens of the sky screen target detection device, the light flux on the photosensitive surface of the optical sensor of the optical lens changes due to the shielding of the target projectile, the changed light flux is extracted, amplified, noise filtered and processed by a level conversion circuit, and finally a fixed pulse width TTL level signal is output to start the high-speed cameras on both sides of the trajectory to collect real-time continuous high-frame frequency video images of the target projectile, the collected images are cached to obtain target projectile sequence image information, and the target projectile sequence image information is sent to the burst point image processing system; the sky screen target trigger device also starts the infrared detection device, and the infrared detection device starts to collect the projectile explosion time information in real time and sends the collected projectile explosion time information to the burst point image processing system.

6. The burst space coordinate measurement method of claim 5, wherein, The step 1) is specifically: Step 1): The burst point image processing system receives the target shell sequence image information sent by the high-speed camera and the shell explosion time information sent by the infrared detection device in real time, and according to the shell explosion time information provided by the infrared detection device, quickly extracts the target shell video information start and end time before and after the target shell explosion in the target shell sequence image information, uses the video frame method to derive each frame of JPEG image I(u e ,v e ,t) with time mark, then quickly extracts the shell image information of the burst point area, and judges the shell image information to determine whether the high-speed camera has captured the image at the explosion time point, and performs the following two operation modes: The first way is: if it is judged that the high-speed camera has taken the image at the explosion time point in the shell image information, the image at the explosion time point is the explosion point image Gd'0(u e ,v e ,t0) of the shell explosion, the image before the shell entering the field of view is taken as the background image M bg (u e ,v e ,t0), the background difference algorithm is used to process the explosion point image Gd'0(u e ,v e ,t0) with time mark, the multi-gray level explosion point image Gd'0(u e ,v e ,t0) after the background processing is binarized, the threshold is determined by using the maximum inter-class variance method, the pixel value greater than or equal to the threshold is set to 1, and the pixel value less than the threshold is set to 0; then the erosion and expansion open operation filtering processing is performed on the binarized explosion point image Gd'0(u e ,v e ,t0) by using the morphological filtering method, the small particle noise is removed, the target boundary is smoothed, the shape and area are not changed, the explosion point target or the shell target and the fire light target is accurately extracted, and the explosion point image information Gd n0 (u e ,v e ,t n0 ) of the shell explosion and the corresponding time T n0 are obtained. Method 2: If it is determined that the high-speed camera did not capture an image at the time of the shell explosion in the shell image information, then the explosion time information obtained by the infrared detection device is used to quickly extract the target shell sequence image information from the shell image information, and the image of the frame before the shell enters the field of view is obtained as the background image M. bg (u e ,v e The target shell sequence image information with time stamps is processed by background subtraction algorithm to reduce background. Then, the multi-grayscale target shell sequence image information after background subtraction is binarized. The threshold is determined by the maximum inter-class variance method. Pixel values ​​greater than or equal to the threshold are set to 1, and pixel values ​​less than the threshold are set to 0. Then, the morphological filtering method is used to perform opening operation filtering on the binarized target shell sequence image information with erosion and dilation to remove small particle noise, smooth the target boundary, and not change its shape and area. The detonation point target or shell target and flash target are accurately extracted to obtain the shell image information Gd of the frame before detonation. n-1 (u e ,v e ,t n-1 ) and the shell flash image information when the first frame after the shell explodes. Gd n1 (u e ,v e ,t n1 ) and the corresponding time T for each frame n-1 T n1 ; Then, the morphologically filtered image of the shell before detonation, Gd, is used. n-1 (u e ,v e ,t n-1 The image of the shell's flash in the first frame after the shell explodes (Gd) n1 (u e ,v e ,t n1 Substituting these values ​​into the frame interpolation algorithm, and combining them with the fireball expansion model of a shell explosion, assuming the weight of the frame before detonation is (L-α) / L, and the weight of the first frame's fire image after detonation is α / L, the intermediate frame image Gd0(u) is obtained by multiplying the two reference frames by their respective weights and then adding them together. e ,v e ,t0); The algorithm is as follows: Wherein, wherein, L is the distance between two original frames, and alpha is the relative distance between the preceding frame and the inserted frame before detonation; The insertion frame time is: T n0 = T n-1 + a In mode two, the generated intermediate frame image is analyzed, and there are two cases: (1) the superposition or partial superposition of the projectile and the fire image, so that the burst point coordinates of the intermediate frame image are the intermediate points of the projectile coordinates and the center pixel coordinates of the fire image; (2) the target projectile is completely overlapped with the fire image, so that the burst point coordinates of the intermediate frame image are the center pixel coordinates of the fire image; Step 2) is specifically: The image Gd'(u e ,v e ,t0) of the explosion point is obtained by detonating the shell obtained in the first way, or the image Gd0(u e ,v e ,t0) of the intermediate frame is obtained by detonating the shell obtained in the second way, and the coordinates (u p0 ,v p0 ) of the explosion point are calculated by using a center-of-mass coordinate extraction algorithm based on a matrix; the quality of each pixel in the area of the explosion point in the image is set as 1, that is, the quality of each pixel is equal to its pixel value, (u e ,v e ) is the coordinate of the pixel of the image, and S is the area of the pixel region, so that the p+q order moment of the target can be represented as: where M is the moment of the image for different values of p, q, f(u e ,v e ) is the mass of a pixel; the zeroth and first moments are calculated separately, and there are three cases as follows: When p=0, q=0, the zero-order moment M(0,0) is obtained: When p=1, q=0, the first-order moment M(0,1) is: When p=0, q=1, the value of the first-order moment M(1,0) is: The center of gravity of the target image can be calculated using the zeroth moment and the first moment, and the bomb point coordinates are represented by (u p0 ,v p0 ). The solution algorithm for the bomb point coordinates is as follows: Wherein, M(1,0) represents the sum of the horizontal coordinates of all pixels of the bullet hole, M(0,1) represents the sum of the column coordinates of all pixels of the bullet hole, and M(0,0) represents the number of pixels contained in the bullet hole; Step 3) is specifically: solving the spatial burst point coordinate model and bringing the burst point coordinates into the model to solve the spatial three-dimensional coordinates of the burst point: Assume that the three-dimensional coordinates of the space burst point P in the world coordinate system are (X P ,Y P ,Z P ), and the image coordinates after imaging are (u P ,v P ); the high-speed camera linear model can be expressed as: The camera distortion model is as follows: wherein, Considering second order radial distortion, distortion coefficients are a1, a2; λ is a scale factor, (u w ,v w ) is the undistorted image coordinate, (R, T) is the extrinsic parameter of the camera, R and T are the rotation matrix and translation vector from the world coordinate system to the camera coordinate system, respectively, and A is the intrinsic parameter matrix of the camera, which can be expressed as: where (u0, v0) is the principal point coordinate of the image coordinate system, f x , f y are the scale factors of the u-axis and the v-axis respectively, and a is the non-perpendicular factor of the u-axis and the v-axis; the high-speed camera calibration needs to refer to 5 parameters f x , f y , a, u0 and v0 of the camera internal parameter matrix, and the distortion coefficients a1 and a2. Before testing, two standard test poles are rotated in multiple directions within the pre-burst point range of the projectile, and the high-speed binocular camera is used to shoot multiple standard test pole moving images in different directions at the pre-burst point of the projectile; according to the imaging relationship of the feature points on the standard test pole on the binocular camera composed of the high-speed cameras on both sides, the five parameters of camera calibration and camera distortion coefficients are calculated; By establishing the image pixel coordinate system, the image coordinate system, the camera coordinate system and the world coordinate system, and combining the conversion relationship between the coordinate systems, the transformation relationship from the pixel coordinate system to the world coordinate system is obtained as follows: wherein Z c is set to 1, s is a coordinate axis inclination parameter, in an ideal case the value is 0, A is a camera internal parameter matrix, an orthogonal rotation matrix R is a cosine combination of the direction of the coordinate axis of the camera coordinate system relative to the world coordinate system, a translation matrix T = [t1t2t3] T is the coordinate of the origin of the camera coordinate system under the world coordinate system; further solving the above formula, the conversion relationship between the system pixel coordinates and the world coordinate system can be obtained as follows: Let The above formula can be simplified to the following spatial burst point coordinate model: [X Y Z 1] T = C -1 [u v 1] T In the formula, (X, Y, Z) is the solved spatial burst point coordinates, and the unit is meter; (u, v) is the pixel coordinate, and the unit is pixel; wherein the blast point coordinates (u p0 ,v p0 ) obtained in step 2) are substituted into the above formula space blast point coordinate model to obtain the space three-dimensional coordinates (X, Y, Z) of the blast point at the blast moment.

Images