CN-121999017-A - Dynamic tracking method, system, equipment and storage medium for explosion fragments

CN121999017ACN 121999017 ACN121999017 ACN 121999017ACN-121999017-A

Abstract

The invention provides a dynamic tracking method, a system, equipment and a storage medium of explosion fragments, which belong to the field of optical testing, and comprise the steps of processing a binocular high-speed image by utilizing Bouguet polar correction algorithm, reducing matching search from two dimensions to one dimension, then providing a Rank-Census mixed binocular matching algorithm, realizing high-precision matching and three-dimensional reconstruction of the same-name fragments in a complex environment by fusing statistical robustness of Rank transformation and texture sensitivity of Census transformation, then introducing a fragment diffusion physical model, defining the explosion distance and diffusion angle range of the fragments according to a charging structure, predicting a candidate space region at the next moment, and finally adopting a speed consistency screening strategy in the preset region, and carrying out track association on fragments with the speed closest to the average value by calculating the arithmetic average value of candidate fragments, thereby realizing stable and continuous tracking of target fragments.

Inventors

WANG CHANGLI
QIAN BINGWEN
WU ZHENGHAO
KE MING
WANG LIANGQUAN
ZHANG BOTIAN
ZHANG XIN

Assignees

西北核技术研究所

Dates

Publication Date: 20260508
Application Date: 20260121

Claims (10)

1. A method for dynamic tracking of an explosion fragment, comprising: Synchronously binocular acquiring high-speed images at continuous moments of an explosion field to form a binocular image sequence, wherein the binocular image sequence comprises a left view and a right view; Selecting a target fragment from the left view of the binocular image sequence based on the current moment, extracting characteristic points of the target fragment as local neighborhood windows, respectively executing Rank transformation and Census transformation on the local neighborhood windows, and fusing the two transformation results to obtain a first composite characteristic; selecting all fragments in the same scanning line with the local neighborhood window of the left view from the right view of the binocular image sequence, selecting the local neighborhood window of each fragment, and sequentially and respectively executing Rank transformation and Census transformation on each local neighborhood window to obtain a plurality of second composite characteristics; The method comprises the steps of determining the space coordinate of a target fragment at the current moment based on the pixel coordinate of the same-name fragment pair in a binocular image sequence, determining the candidate space region of the target fragment in the binocular image sequence at the next moment according to the space coordinate of the target fragment at the current moment and a preset fragment motion constraint range, calculating the space coordinate of each fragment in the candidate space region at the front and rear moment, determining the instantaneous speed of each fragment according to the coordinates of the front and rear moment, determining the target fragment at each moment according to the instantaneous speed, and completing the dynamic tracking of the target fragment according to the space coordinate of the target fragment at different moments.
2. The dynamic tracking method of explosion fragments according to claim 1, wherein the Rank transformation comprises the steps of calculating the number of pixel points with gray values smaller than those of central pixels in a local neighborhood window to be used as local statistical features, generating binary features by calculating gray differences between neighborhood pixels and central pixels in the local neighborhood window, and fusing the local statistical features with the binary features to obtain first composite features.
3. The dynamic tracking method of the explosion fragments according to claim 1 is characterized in that the method is characterized in that based on the current moment, a target fragment is selected from the left view of a binocular image sequence, before characteristic points of the target fragment are extracted as local neighborhood windows, a Bouguet polar correction algorithm is adopted to correct the binocular image sequence, and specifically the method comprises the steps of obtaining relative angle data and relative distance data of a left camera and a right camera of the binocular image sequence acquisition device, determining a rotation matrix R and a translation vector T of the left camera and the right camera by utilizing the relative angle data and the relative distance data, and correcting the binocular high-speed image sequence according to the rotation matrix R and the translation vector T to obtain a corrected image pair sequence with aligned lines.
4. The method of claim 1, wherein an absolute difference of Rank changes between the first composite feature and the second composite feature is calculated, a hamming distance of Census changes between the first composite feature and the second composite feature is calculated, and a difference value between the first composite feature and each second composite feature is determined from the absolute difference of Rank changes and the hamming distance of Census changes.
5. The method for dynamically tracking explosion fragments according to claim 1, wherein the calculating the space coordinates of the front and rear time points of each fragment in the candidate space region, determining the instantaneous speed of each fragment from the coordinates of the front and rear time points, and determining the target fragment at each time point from the instantaneous speed comprises: The method comprises the steps of determining the instantaneous speed of each fragment by utilizing the space coordinates of the front moment and the rear moment, determining the average speed of the whole explosion field by the instantaneous speed of each fragment, and selecting the fragment corresponding to the instantaneous speed with the smallest difference with the average speed as the target fragment of the next moment.
6. The method of claim 1, wherein the predetermined fragment motion constraint range is determined based on physical structural parameters of the explosive charge and a fragment diffusion model.
7. The method of claim 1, wherein determining candidate spatial regions in the binocular image sequence of the target fragment at a next moment further comprises introducing a dynamic compensation factor to adaptively adjust a diffusion angle range of each fragment to compensate for a trajectory deviation caused by aerodynamic disturbance.
8. A dynamic tracking method system for explosive fragments, comprising: the data acquisition module is used for synchronously and binocular acquiring high-speed images at continuous moments of an explosion field to form a binocular image sequence, wherein the binocular image sequence comprises a left view and a right view; The object searching module is used for selecting a target fragment from the left view of the binocular image sequence based on the current moment, extracting characteristic points of the target fragment as local neighborhood windows, respectively executing Rank transformation and Census transformation on the local neighborhood windows, and fusing the two transformation results to obtain a first composite characteristic; selecting all fragments in the same scanning line with the local neighborhood window of the left view from the right view of the binocular image sequence, selecting the local neighborhood window of each fragment, and sequentially and respectively executing Rank transformation and Census transformation on each local neighborhood window to obtain a plurality of second composite characteristics; The positioning module is used for determining the space coordinate of the target fragment at the current moment based on the pixel coordinates of the same-name fragment pair in the binocular image sequence, determining the candidate space region of the target fragment in the binocular image sequence at the next moment according to the space coordinate of the target fragment at the current moment and the preset fragment motion constraint range, calculating the space coordinates of each fragment in the candidate space region at the front and rear moments, determining the instantaneous speed of each fragment according to the coordinates of the front and rear moments, determining the target fragment at each moment according to the instantaneous speed, and completing the dynamic tracking of the target fragment according to the space coordinates of the target fragment at different moments.
9. A computer device comprising a memory, a processor and a computer program stored on the memory, characterized in that the processor executes the computer program to carry out the steps of the method according to any one of claims 1 to 7.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when loaded by a processor, is able to carry out the steps of the method according to any one of claims 1 to 7.

Description

Dynamic tracking method, system, equipment and storage medium for explosion fragments Technical Field The invention belongs to the field of optical testing, and particularly relates to a dynamic tracking method, a system, equipment and a storage medium for explosion fragments. Background The explosion fragments are core carriers of the warhead damage effect, have the characteristics of spatial distribution, movement track and speed attenuation, and have important significance for weapon power assessment, protection structure design and target damage mechanism research. The accurate, stable and full-field dynamic tracking and measurement of the explosion fragments are realized, and the method is a key technical requirement in the field. At present, the explosion fragment tracking method based on high-speed imaging is a matching and reconstruction method based on stereoscopic vision. According to the method, the three-dimensional positions of the fragments are restored through feature matching of the left image and the right image, and then the motion trail of the fragments is tracked. When the stereo vision-based matching and reconstruction method is applied to explosion fragment tracking, two outstanding bottlenecks are faced, namely, the fragment target is small in size, weak in texture and similar in shape, the characteristic distinction degree is low under complex interference, so that the matching precision is poor, the mismatching rate is high, and the two-dimensional image space is required to be searched for in order to find corresponding points in left and right images, the calculation complexity is high, and the real-time requirement of high-speed fragment tracking is difficult to meet. Therefore, in a high-interference and high-density explosion fragment field, the aim of simultaneously realizing high-efficiency and high-precision binocular stereo matching has become a key difficult problem for restricting explosion fragment tracking. Disclosure of Invention In order to solve the background problem, the invention provides a dynamic tracking method, a system, equipment and a storage medium for explosion fragments. In order to achieve the above object, the present invention provides a method for dynamically tracking an explosion fragment, including: and synchronously binocular acquiring high-speed images at continuous moments of the explosion field to form a binocular image sequence, wherein the binocular image sequence comprises a left view and a right view. Selecting a target fragment from the left view of the binocular image sequence based on the current moment, extracting characteristic points of the target fragment as local neighborhood windows, respectively executing Rank transformation and Census transformation on the local neighborhood windows, fusing two transformation results to obtain a first composite characteristic, screening all fragments in the same scanning line with the local neighborhood windows of the left view from the right view of the binocular image sequence, selecting the local neighborhood windows of all fragments, sequentially respectively executing Rank transformation and Census transformation on all the local neighborhood windows to obtain a plurality of second composite characteristics, calculating the difference value of the first composite characteristic and each second composite characteristic, and selecting fragments corresponding to the minimum difference value from all the difference values to form a same-name fragment pair with the target fragment. The method comprises the steps of determining the space coordinate of a target fragment at the current moment based on the pixel coordinate of the same-name fragment pair in a binocular image sequence, determining the candidate space region of the target fragment in the binocular image sequence at the next moment according to the space coordinate of the target fragment at the current moment and a preset fragment motion constraint range, calculating the space coordinate of each fragment in the candidate space region at the front and rear moment, determining the instantaneous speed of each fragment according to the coordinates of the front and rear moment, determining the target fragment at each moment according to the instantaneous speed, and completing the dynamic tracking of the target fragment according to the space coordinate of the target fragment at different moments. Preferably, the Rank transformation comprises the steps of calculating the number of pixel points with gray values smaller than those of a central pixel in a local neighborhood window to serve as local statistical features, wherein the Census transformation comprises the steps of generating binary features by calculating gray differences between the neighborhood pixels in the local neighborhood window and the central pixel, and fusing the local statistical features with the binary features to obtain first composite features. The method comprises the s