A particle trajectory tracking method, apparatus, device, and medium based on the Retina algorithm

Abstract

This invention relates to a particle trajectory tracking method based on the Retina algorithm. The method includes: establishing a retina parameter space; dividing the established retina parameter space into discrete subspaces, each discrete subspace being regarded as a retina computing unit; the retina computing units forming a retina computing array; inputting sample data into the retina computing units; obtaining the similarity weight and weight value of each element in the sample data within the retina computing units; obtaining the weight value distribution of the input samples in the retina parameter space based on the weight values ​​of all retina computing units; and establishing a tracking trajectory based on the weight value distribution and the discrete subspace of the retina computing unit corresponding to the maximum weight value. This enables rapid localization and reconstruction of high-energy particle trajectories within a specified spatial range, providing a solution for radiation localization in space and particle trajectory imaging in high-energy physics experiments.

Description

Technical Field

[0001] This invention relates to the field of tracking, and more particularly to a particle trajectory tracking method based on the Retina algorithm. Background Technology

[0002] As the collision energies in high-energy physics experiments increase, the data flow generated by these experiments increases exponentially. The demand for real-time online trajectory tracking in high-energy physics experiments is growing. It can extract the trajectory feature information of particles of interest in a targeted and rapid manner while reducing the data transmission bandwidth of the DAQ (Data Acquisition System).

[0003] Research on real-time tracking algorithms began quite early. In the LHCb physics detection experiment, Günther et al. solved key parts of "forward tracking," such as using SIMD instructions in Hough transform and cluster search, achieving partial event reconstruction, track reconstruction, tracking efficiency, false track rate, and high throughput. Hennequin, A. et al. proposed a parallel reconstruction algorithm for the vertex detector in the LHCb experiment. The efficiency and accuracy of tracking particle straight-line trajectories need further improvement. Summary of the Invention

[0004] This invention addresses the technical problems existing in the prior art by providing a particle trajectory tracking method based on the Retina algorithm, which enables rapid localization and reconstruction of high-energy particle trajectories within a specified spatial range, providing a solution for radiation localization in space and particle trajectory imaging in high-energy physics experiments.

[0005] According to a first aspect of the present invention, a particle trajectory tracking method based on the Retina algorithm is provided, comprising:

[0006] Step 1: Divide the two-dimensional space into spatial units according to the set granularity, and use the spatial units as cell units in the retinal algorithm to establish a cell array;

[0007] Step 2: Use detectors to obtain the coordinate position information of pixels when a particle passes through the detector planes of each layer;

[0008] Step 3: Calculate the similarity weights of each coordinate position information of the particle and each cell unit based on the vertical distance. Sum the similarity weights of each coordinate position information of a cell unit and the particle to obtain the similarity weight value of the cell unit. Traverse all cell units in the cell array, determine the position of the cell unit corresponding to the largest similarity weight value, determine the straight trajectory of the particle based on the position, and output it.

[0009] According to a first aspect of the present invention, a particle trajectory tracking method based on the Retina algorithm is provided, comprising:

[0010] Step 1': Divide the two-dimensional space into spatial units according to the set granularity, and use the spatial units as super retinal units in the retinal algorithm to establish a cell array;

[0011] Step 2': Use detectors to obtain the coordinate position information of pixels when multiple particles pass through the detector planes of each layer;

[0012] Step 3': Calculate the similarity weights of each coordinate position information of the particle and each super retina unit based on the vertical distance. Sum the similarity weights of each coordinate position information of a super retina unit and the particle to obtain the similarity weight value of the super retina unit. Filter out super retina units whose similarity weight values ​​exceed the set threshold sum_th1.

[0013] Step 4': Divide each of the selected super retinal units into multiple cell units, calculate the similarity weight value of the cell unit, and select cell units whose similarity weight value exceeds the set threshold sum_th2; determine the straight trajectory of the particle based on the position and output it.

[0014] According to a third aspect of the present invention, a particle trajectory tracking device based on the Retina algorithm is provided, comprising:

[0015] Cell array setup unit, detector, and trajectory output unit;

[0016] The cell array establishment unit is used to divide the two-dimensional space into various spatial units according to a set granularity, and use the spatial units as cell units in the retinal algorithm to establish a cell array.

[0017] The detector is used to detect and acquire the coordinate position information of the pixel points when a particle passes through the plane of each layer of detectors;

[0018] The trajectory output unit is used to calculate the similarity weight between the coordinate position information of each particle and each cell unit based on the longitudinal distance, sum the similarity weights of each coordinate position information of a cell unit and the particle to obtain the similarity weight value of the cell unit, traverse all cell units in the cell array, determine the position of the cell unit corresponding to the largest similarity weight value, determine the straight trajectory of the particle based on the position and output it.

[0019] According to a fourth aspect of the present invention, an electronic device is provided, including a memory and a processor, wherein the processor is configured to implement a particle trajectory tracking method based on the Retina algorithm when executing a computer management program stored in the memory.

[0020] According to a fifth aspect of the present invention, a computer-readable storage medium is provided, on which a computer management class program is stored, wherein when executed by a processor, the computer management class program implements the steps of a particle trajectory tracking method based on the Retina algorithm.

[0021] This invention provides a particle trajectory tracking method, device, electronic device, and storage medium based on the Retina algorithm. It proposes a retina algorithm and an upgraded iterative retina algorithm to track single-track and multi-track feature information in a single event sample. The invention also conducts position resolution, efficiency, and purity tests. The final resolution of single-track tracking reaches within 40 μm, which is less than 50% of the size of a single pixel (83 μm). The position resolution of multi-track tracking is comparable to that of single-track tracking, while the tracking efficiency and purity are higher than 97%. Attached Figure Description

[0022] Figure 1 A flowchart illustrating a particle trajectory tracking method based on the Retina algorithm provided by this invention;

[0023] Figure 2 This is a schematic diagram of an embodiment of the Retina algorithm provided by the present invention for tracing a straight line in a two-dimensional plane;

[0024] Figure 3 This is a schematic diagram of the calculation and distribution of retinal weights in the Retina algorithm provided by the present invention;

[0025] Figure 4 This is a flowchart of multi-particle track finding using the iterative retinal algorithm provided in this embodiment of the invention;

[0026] Figure 5 A two-dimensional map of multi-track formation provided in an embodiment of the present invention;

[0027] Figure 6 A schematic diagram of the iterative retina algorithm provided in this embodiment of the invention;

[0028] Figure 7 A flowchart illustrating the statistical count of perfectly matched tracks, over-matched tracks, and unmatched tracks in the iterative retina algorithm provided in this embodiment of the invention.

[0029] Figure 8 This is a flowchart illustrating the statistical analysis of the number of false tracks in the iterative retina algorithm provided in this embodiment of the invention.

[0030] Figure 9 A schematic diagram of the hardware structure of a possible electronic device provided by the present invention;

[0031] Figure 10 This is a schematic diagram of the hardware structure of a possible computer-readable storage medium provided by the present invention. Detailed Implementation

[0032] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0033] Example 1

[0034] Embodiment 1 provided by the present invention is an embodiment of a particle trajectory tracking method based on the Retina algorithm provided by the present invention. In this embodiment, single particle trajectory tracking is performed on single particle trajectory event samples. Figure 1 A flowchart of a particle trajectory tracking method based on the Retina algorithm provided in this embodiment of the invention is shown below. Figure 1 As shown, the embodiment includes:

[0035] Step 1: Divide the two-dimensional space into spatial units according to the set granularity, and use the spatial units as cell units in the retinal algorithm to establish a cell array.

[0036] The retina algorithm is a pattern recognition algorithm inspired by the visual characteristics of biological retinas and invented by imitating the biological visual system. This invention uses the retina algorithm to recognize straight line patterns on a two-dimensional plane. First, the two-dimensional space is divided into several small spatial units according to a pre-defined granularity. Each small spatial unit is called a cell unit (Retina cell) of the algorithm, and the corresponding two-dimensional parameters are (i, j).

[0037] The specific partitioning process includes: sequentially setting the boundary conditions of the cell array in the horizontal and vertical directions, i.e., the value range of the cell array on the horizontal and vertical axes; dividing the array into i*j small rectangular Retina cells along the horizontal and vertical axes, i.e., dividing the array into cell computing units of i rows * j columns, with the total number of cell units being the granularity set by the algorithm. The horizontal dimension can be represented as... :{ , ,... The vertical dimension can be represented as :{ , ,..., }

[0038] Step 2: Use the detector to obtain the coordinate position information of the pixel point when a particle passes through the detector plane of each layer.

[0039] In single-particle track finding, simulation experiments utilize high-energy heavy-ion Kr (average energy 25 MeV / µ) rays passing through a gas chamber composed of Ar and CO2 (Ar:CO2=70:30). As the charged particle passes through the gas detector, it interacts electromagnetically with electrons in the gas atoms, leading to ionization and excitation of the gas medium. The energy lost during this interaction is converted into low-energy photons and electron-ion pairs. These electron-ion pairs form a straight trajectory along the particle's flight direction, landing perpendicularly on the surface of the Topmetal pixel detector chip under an applied electric field. The Topmetal chip collects the electrons and generates a straight trajectory sample within an effective two-dimensional region.

[0040] In one possible embodiment, electrons are drifted onto the pixel detector by applying an electric field (300 V / cm) perpendicular to the plane of the pixel detector within the gas chamber. Finally, the Topmetal chip collects these charges, forming linear particle trajectories in the detection region, which serve as input samples for the algorithm.

[0041] Step 3: Calculate the similarity weights of each coordinate position information of the particle and each cell unit based on the vertical distance. Sum the similarity weights of each coordinate position information of a cell unit and the particle to obtain the similarity weight value of the cell unit. Traverse all cell units in the cell array, determine the position of the cell unit corresponding to the largest similarity weight value, determine the straight trajectory of the particle based on the position, and output it.

[0042] In the specific implementation process, the case sample information is sequentially input into the Retina computing array, and similarity weights are calculated for each case sample and each Retina cell. As can be seen from the calculation process of the Retina algorithm, the calculation of the similarity weights of each computing unit in the Retina computing array can be performed simultaneously. This makes the algorithm calculation process highly parallel and has a short computational delay, giving it the potential to be a real-time tracking algorithm.

[0043] In one possible embodiment, the formula (1) for obtaining the similarity weight value of cell units based on the longitudinal distance is:

[0044]

[0045] Where i and j represent the horizontal and vertical serial numbers of the cell unit / super retinal unit, respectively; m and k represent the total number and serial number of the layers of the detector's parallel plane, respectively. The similarity weight value is the coordinate position information of the cell unit / super retina unit in the i-th row and j-th column and the particle detected by the detector in the k-th layer. This represents the coordinate position information of the particles detected by the k-th layer detector and the distance of the cell unit / super retina unit in the i-th row and j-th column along the vertical axis. This is the adjustment factor for the operation, and its size is determined by the granularity of the retina.

[0046] In one possible embodiment, the process of determining and outputting the linear trajectory of the particle based on the position includes:

[0047] Determine the horizontal and vertical serial numbers i and j of the cell unit / super retina unit, and use the slope and intercept of the cell unit / super retina unit in the i-th row and j-th column as the slope and intercept of the particle linear trajectory.

[0048] like Figure 2 and Figure 3 The figures shown are schematic diagrams illustrating the tracing of a straight line in a two-dimensional plane, the calculation of retinal weights, and the distribution of weight values ​​in the Retina algorithm provided in this embodiment of the invention. A straight line in two-dimensional space is represented as y = kx + b in a rectangular coordinate system. The two-dimensional parameter space of the Retina algorithm is defined as (k, b), where the horizontal axis represents the slope k of the predicted line, with a value range limited to -1 to 1; the vertical axis represents the ordinate b of the predicted line, with a value range of -1 to 1.

[0049] Figure 2 In the given embodiment, the scan granularity of the Retina algorithm is set to i*j=4*4=16. The granularity is set to the Retina cell computational unit in the entire parameter space. ~ ,in This represents the Retina cell computing unit in the i-th row and j-th column.

[0050] The corresponding line Lij is defined as the Retina pre-matched line, and its corresponding line expression is y = jx + i. For example... Figure 2 The left side shows the "slope-intercept" two-dimensional parameter space. The straight lines in the "xy" two-dimensional physical space corresponding to cell23 and cell41 in the two-dimensional parameter space are respectively... Figure 2 L23 and L41 are shown on the right.

[0051] The actual sample straight line trajectory is Figure 3 The coordinates of the straight line track information shown on the left, composed of gray dots, are represented by the vertical distance between the coordinate points and the Retina-predicted straight line L23. , The similarity weights and sum23 of the sample in cell23 can be calculated using formula (1). Similarly, the similarity weights of the sample in each cell unit of the cell array can be calculated, thus obtaining the weights and distribution of the sample in the Retina cell array. ~ like Figure 3 As shown on the right. Compare the sizes and select the maximum value. The position coordinates (i, j) in the parameter space represent the slope and intercept parameters of the particle's straight track, as identified and reconstructed by the algorithm. At this point, the Retina algorithm has completed the tracking and reconstruction of a straight track.

[0052] In the simulation verification, the number of line samples generated each time was set to 2000, and the range of parameters k and b was set to 0-10. The granularity of the Retina algorithm was set to four cases from high to low: 10*10, 20*20, 49*49, and 99*99. The slope resolution and y-intercept resolution of the line trajectories identified by the algorithm under the four different algorithm granularities are shown in Table 1.

[0053] Table 1

[0054]

[0055] The slope and intercept resolutions reconstructed by the Retina algorithm in the experimental results are given in the form of standard deviation. Table 1 shows that the accuracy of the algorithm's tracking increases with the granularity of the Retina array. When the algorithm granularity increases to 99*99, the slope resolution and intercept resolution reach 0.03 and 0.14 respectively, which are quite high, verifying the feasibility of the Retina algorithm for tracking straight line feature information.

[0056] This invention provides a particle trajectory tracking method based on the Retina algorithm. By combining algorithm research with conventional pixel detector chips, it enables rapid localization and reconstruction of high-energy particle trajectories within a specified spatial range, providing a solution for radiation localization in space and particle trajectory imaging in high-energy physics experiments.

[0057] Example 2

[0058] Embodiment 2 of this invention is an embodiment of a particle trajectory tracking method based on the Retina algorithm provided by this invention. In this embodiment, the IR (Iterative-Retina) algorithm is used for multi-particle trajectory tracking of particle multi-track event samples, such as... Figure 4The diagram shows a flowchart of the iterative retina algorithm for multi-particle track finding provided in an embodiment of the present invention. Figure 1 and Figure 4 It can be seen that embodiments of this trajectory tracking method include:

[0059] Step 1': Divide the two-dimensional space into spatial units according to the set granularity, and use the spatial units as super retinal units in the retinal algorithm to establish a cell array.

[0060] Step 2': Use detectors to obtain the coordinate position information of pixels when multiple particles pass through the detector planes of each layer.

[0061] When performing multi-particle track tracking, multi-track samples are used as input to the IR algorithm. Under the same conditions and environment as single-particle sample generation, multiple Kr heavy ion rays are input during the generation of a single sample event. In this way, Topmetal can acquire samples formed by multiple straight tracks each time. Figure 5 As shown, this sample information is used as a multi-particle track sample for algorithm tracking input. During the simulation, the number of incident Kr heavy ions is set to 3 to 7.

[0062] Step 3': Calculate the similarity weights of each coordinate position information of the particle and each super retina unit based on the longitudinal distance. Sum the similarity weights of each coordinate position information of a super retina unit and the particle to obtain the similarity weight value of the super retina unit. Filter out super retina units whose similarity weight values ​​exceed the set threshold sum_th1.

[0063] Step 4': Divide each of the selected super retinal units into multiple cell units, calculate the similarity weight value of the cell unit, and select cell units whose similarity weight value exceeds the set threshold sum_th2; determine the straight trajectory of the particle based on the position and output it.

[0064] The working principle of the IR algorithm is as follows: Figure 6 As shown, combined with Figure 6 It can be seen that the IR algorithm calculates the target parameter space through multiple iterations of the retina algorithm, thereby finding the maximum value in multiple regions as the target trajectory to be identified.

[0065] In this embodiment of the invention, taking the two-iteration IR algorithm as an example, the granularity is set to N*N in the first retina calculation, and the entire parameter space is divided into i*j super cells. In the second retina calculation, the granularity is set to M*M, and the parameter space contained in each selected super cell is further divided into N*M cells. The pixel threshold and distance threshold settings in the two retina calculations are the same as those in the retina algorithm in Embodiment 1. However, the filtering thresholds sum_th1 and sum_th2, which need to be set twice, are used to filter out the Retinacell cells containing the maximum values ​​to complete the tracking task.

[0066] The sum value of the sample in each supercell can be obtained through the first retinal algorithm scan. Supercells with sum values ​​exceeding the threshold sum_th1 are selected (e.g., ...). Figure 6 The yellow square represents the super cell, and the algorithm then proceeds to the second retinal algorithm for iterative calculation.

[0067] In the second iteration, the cell containing the maximum sum value in each Supercell is identified, and it is determined whether the sum value corresponding to that cell exceeds a set threshold sum_th2. If it exceeds the threshold sum_th2, a pattern is successfully identified in that cell space, such as... Figure 6 The cell region containing the green square. If the threshold sum_th2 is not exceeded, then no valid pattern was matched in this super cell.

[0068] Compared to the Retina algorithm, the IR algorithm reduces the overall computational cost while maintaining the same scan granularity. Assuming the IR algorithm selects H supercell units in the first retinal calculation, the total number of retinal calculation units traversed by the algorithm is N*N + H*M*M. In contrast, the Retina algorithm requires N retinal calculation units to achieve the same scan granularity. 2 *M 2 The computational cost of the IR algorithm is (1 / M) that of the conventional Retina algorithm. 2 +H / N 2 () times, a significant decrease.

[0069] In one possible embodiment, the formula for obtaining the similarity weight value of cell unit / super retinal unit based on longitudinal distance is:

[0070]

[0071] Where i and j represent the horizontal and vertical serial numbers of the cell unit / super retinal unit, respectively; m and k represent the total number and serial number of the layers of the detector's parallel plane, respectively. The similarity weight value is the coordinate position information of the cell unit / super retina unit in the i-th row and j-th column and the particle detected by the detector in the k-th layer. This represents the coordinate position information of the particles detected by the k-th layer detector and the distance of the cell unit / super retina unit in the i-th row and j-th column along the vertical axis. This is the adjustment factor for the operation, and its size is determined by the granularity of the retina.

[0072] In one possible embodiment, the process of determining and outputting the linear trajectory of the particle based on the position includes:

[0073] Determine the transverse and longitudinal index numbers i and j of the cell unit / super retinal unit, and use i and j as the slope and intercept of the particle's linear trajectory.

[0074] For multipath tracking algorithms, efficiency and purity are two important metrics for evaluating their performance. They reflect the algorithm's ability to efficiently identify as many straight-line trajectories as possible from all samples. In one possible implementation, the trajectory tracking method further includes:

[0075] Multiple trajectory sample information is obtained by using a detector. The trajectory sample information is the coordinate position information of the pixel points when multiple particles pass through the plane of each detector layer.

[0076] After using the IR algorithm to trace N multi-track samples and find M straight-line trajectories, the result analysis requires classifying the M trajectories one by one according to the defined matching patterns. The classification of matching patterns includes:

[0077] If there is exactly one straight line trajectory that matches a track in the standard library, it is considered a perfect matching track.

[0078] If at least two straight-line trajectories are identical to one track in the standard library, it is determined to be an over-perfect matching track.

[0079] If any track in the standard track library is different from all the straight track tracks, it is judged as an unmatched track.

[0080] If a straight track is not identical to any track in the standard track library, it is considered a false track.

[0081] like Figure 7 The diagram shown is a flowchart illustrating the statistical count of perfectly matched tracks, over-matched tracks, and unmatched tracks in the iterative retina algorithm provided in this embodiment of the invention. Figure 7It can be seen that the original trajectories in the samples are used as input, and similarity matching is performed sequentially with all trajectories found by the IR algorithm. For each successful match, the matching count F increases by one. The similarity matching criterion is to count the number of identical coordinate points Num_s between the original trajectory and the trajectory identified by the IR algorithm. A match is successful when this number exceeds a threshold Num_th; otherwise, the match fails. Finally, after all the original trajectories in the samples have been statistically analyzed, the straight-line trajectory with a matching count F equal to 1 is the perfectly matched trajectory. Straight-line trajectories with a matching count F greater than 1 are overmatched trajectories, and straight-line trajectories with a matching count F equal to 0 are unmatched trajectories.

[0082] In one possible embodiment, such as Figure 8 The diagram shown is a flowchart illustrating the statistical analysis of the number of false tracks in the iterative retina algorithm provided in this embodiment of the invention. Figure 8 As can be seen, in the process of counting fake tracks, the trajectory found by the IR algorithm is used as input, and similarity matching is performed with all trajectories in the sample in turn. If no match is found, the number of fake tracks N_fake is incremented by one. Finally, after inputting all the trajectories reconstructed by the IR algorithm, the number of fake track patterns can be obtained.

[0083] The efficiency of the trajectory tracking method is calculated based on the number of four matching trajectories. and purity The calculation formula is as follows:

[0084] ;

[0085] .

[0086] Where N_repeat, N_miss, N_true, and N_fake represent the number of perfectly matched tracks, overmatched tracks, unmatched tracks, and fake tracks, respectively. This represents the total number of straight line trajectories that originally existed in the sample.

[0087] This invention presents two algorithms based on the retina algorithm for tracking single-track and multi-track samples: the original retina algorithm and the iterative retina algorithm. The performance of the algorithms is evaluated by repeatedly experimenting with appropriate algorithm granularity and other parameters to assess the reconstruction accuracy and tracking efficiency of the algorithms on the experimental samples.

[0088] The original retina algorithm achieves a positional resolution within 40μm, less than 50% of the size of a single pixel (83um*83um). The iterative retina algorithm, with comparable positional and slope resolution to the Retina algorithm, achieves tracking efficiency and purity exceeding 97%, enabling efficient and accurate linear particle trajectory tracking. The final results confirm that the retina algorithm and the iterative retina algorithm can successfully perform linear particle trajectory tracking on the Topmetal pixel detector chip.

[0089] This invention proposes a particle trajectory tracking method based on the Retina algorithm, enabling fast real-time particle trajectory tracking of straight particle tracks detected by a pixel detector in a two-dimensional planar space. Specifically: 1. A particle trajectory tracking algorithm based on the Retina algorithm is designed. This algorithm achieves tracking of a single Kr ion's straight track within a two-dimensional space covered by a Topmetal chip with a 72*72 pixel array, achieving a position resolution within 40µm, less than 50% of the size of a single pixel (83µm*83µm); 2. Based on the Retina algorithm, it is further extended and upgraded to an iterative Retina algorithm. This algorithm achieves simultaneous tracking of multiple particle trajectories in a single event through iterative Retina calculations. While maintaining a position resolution comparable to single-particle track tracking, the algorithm's tracking efficiency and purity are both better than 97%. Furthermore, compared to previous algorithms, this algorithm has very high parallelism, consumes fewer resources, and is easier to implement in real-time, online track tracking applications on a FPGA.

[0090] Example 3

[0091] Embodiment 3 of this invention is an example of determining the parameters of the retina algorithm in a particle trajectory tracking method based on the Retina algorithm provided by this invention. The two-dimensional parameter space of the retina algorithm is (k, b). Here, k represents the slope of the line, and its value range is set to -1 to 1. And b represents the y-intercept of the line, and its value range is set to 0 to 72 pixels.

[0092] To improve the computational accuracy of the retina algorithm in this experimental environment, three new influencing factors were added to the step of calculating the weight value W in the retina algorithm computation unit: pixel value, pixel threshold, and distance threshold.

[0093] In one possible embodiment, the pixel value An of each input sample coordinate point n is added as an influence factor to the similarity weight calculation formula (1). The calculation formula becomes formula (2) as follows:

[0094]

[0095] Using formula (2) allows coordinates with higher brightness (pixel values) in the trajectory samples to have a higher matching weight in the predicted line L represented by the corresponding view cell. This can improve the tracking accuracy of the retina algorithm.

[0096] In one possible embodiment, a distance threshold d_th and a pixel threshold A_th are added to the similarity weight calculation formula (2). The calculation formula becomes formula (3) as follows:

[0097]

[0098] Since a straight track sample contains a large number of valid coordinate points, formula (3) sets two thresholds d_th and A_th to allow the algorithm to ignore the contribution of coordinate positions that are far from the predicted straight line L (dij_n>d_th) or have low pixel values ​​An to the weight value W, thereby reducing the computational load of the algorithm while improving the computational efficiency and tracking accuracy of the algorithm.

[0099] Based on the generated heavy ion trajectory samples, the pixel threshold value A_th is set to 10 brightness levels, while the distance threshold value d_th is set to 1.5 times the size of a single retinal computing unit on the longitudinal intercept parameter axis, i.e., 1.5*83*72 / J um.

[0100] Example 4

[0101] Embodiment 4 of the present invention is an embodiment of a particle trajectory tracking device based on the Retina algorithm provided by the present invention. The embodiment of the trajectory tracking device includes: a cell array establishment unit, a detector, and a trajectory output unit.

[0102] The cell array establishment unit is used to divide the two-dimensional space into various spatial units according to a set granularity, and use the spatial units as cell units in the retina algorithm to establish the cell array.

[0103] The detector is used to detect and obtain the coordinate position information of the pixels when a particle passes through the detector planes of each layer.

[0104] The trajectory output unit is used to calculate the similarity weight between the coordinate position information of each particle and each cell unit based on the longitudinal distance. The similarity weight between a cell unit and the coordinate position information of each particle is summed to obtain the similarity weight value of the cell unit. The unit traverses all cell units in the cell array, determines the position of the cell unit corresponding to the largest similarity weight value, determines the straight trajectory of the particle based on the position, and outputs it.

[0105] It is understood that the particle trajectory tracking device based on the Retina algorithm provided by the present invention corresponds to the particle trajectory tracking method based on the Retina algorithm provided in the foregoing embodiments. The relevant technical features of the particle trajectory tracking device based on the Retina algorithm can be referred to the relevant technical features of the particle trajectory tracking method based on the Retina algorithm, and will not be repeated here.

[0106] Please see Figure 9 , Figure 9 A schematic diagram illustrating an embodiment of the electronic device provided in this invention. For example... Figure 9 As shown, this embodiment of the invention provides an electronic device, including a memory 1310, a processor 1320, and a computer program 1311 stored in the memory 1310 and executable on the processor 1320. When the processor 1320 executes the computer program 1311, it performs the following steps: dividing a two-dimensional space into spatial units according to a set granularity, and establishing a cell array using the spatial units as cell units in the retinal algorithm; using a detector to detect and obtain the coordinate position information of a pixel point when a particle passes through the plane of each layer of detectors; calculating the similarity weight between the coordinate position information of the particle and each cell unit based on the longitudinal distance; summing the similarity weights of a cell unit and the coordinate position information of the particle to obtain the similarity weight value of the cell unit; traversing all cell units in the cell array, determining the position of the cell unit corresponding to the largest similarity weight value; determining the straight-line trajectory of the particle based on the position and outputting it.

[0107] Please see Figure 10 , Figure 10 This is a schematic diagram illustrating an embodiment of a computer-readable storage medium provided by the present invention. (See diagram below.) Figure 10 As shown, this embodiment provides a computer-readable storage medium 1400, on which a computer program 1411 is stored. When the computer program 1411 is executed by a processor, it performs the following steps: dividing the two-dimensional space into spatial units according to a set granularity, and establishing a cell array using the spatial units as cell units in the retinal algorithm; using a detector to detect and obtain the coordinate position information of the pixel points when a particle passes through the planes of each layer of detectors; calculating the similarity weight between the coordinate position information of the particle and each cell unit based on the longitudinal distance; summing the similarity weight between a cell unit and the coordinate position information of the particle to obtain the similarity weight value of the cell unit; traversing all cell units in the cell array, determining the position of the cell unit corresponding to the largest similarity weight value; determining the straight-line trajectory of the particle based on the position and outputting it.

[0108] This invention provides a particle trajectory tracking method, device, electronic device, and storage medium based on the Retina algorithm. It proposes a retina algorithm and an upgraded iterative retina algorithm to track single-track and multi-track feature information in a single event sample. The method performs position resolution, efficiency, and purity tests. The final resolution of single-track tracking reaches less than 40 μm, which is less than 50% of the size of a single pixel (83 μm). The position resolution of multi-track tracking is comparable to that of single-track tracking, while the tracking efficiency and purity are higher than 97%.

[0109] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0110] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0111] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0112] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0113] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0114] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0115] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A particle trajectory tracking method based on the Retina algorithm, characterized in that, The trajectory tracking method includes: Step 1: Divide the two-dimensional space into spatial units according to the set granularity, and use the spatial units as cell units in the retinal algorithm to establish a cell array; Step 2: Use detectors to obtain the coordinate position information of pixels when a particle passes through the detector planes of each layer; Step 3: Calculate the similarity weight between the coordinate position information of each particle and each cell unit based on the vertical distance. Sum the similarity weights between a cell unit and the coordinate position information of each particle to obtain the similarity weight value of the cell unit. Traverse all cell units in the cell array to determine the position of the cell unit corresponding to the largest similarity weight value. Determine the straight-line trajectory of the particle based on the position of the cell unit and output it. The formula for obtaining the similarity weight value of cell units based on vertical distance is: Where i and j represent the horizontal and vertical serial numbers of the cell unit, respectively; m and n represent the total number and serial number of the layers in the parallel plane of the detector, respectively. The similarity weight value is the coordinate position information of the cell unit in the i-th row and j-th column and the particle detected by the n-th layer detector. This represents the coordinate position information of the particle detected by the nth layer detector and the distance of the cell unit in the i-th row and j-th column along the vertical axis. This is the adjustment factor for the operation; This represents the pixel value of the particle detected by the nth layer detector.

2. A particle trajectory tracking method based on the Retina algorithm, characterized in that, The trajectory tracking method includes: Step 1': Divide the two-dimensional space into spatial units according to the set granularity, and use the spatial units as super retinal units in the retinal algorithm to establish a cell array; Step 2': Use detectors to obtain the coordinate position information of pixels when multiple particles pass through the detector planes of each layer; Step 3': Calculate the similarity weights of each coordinate position information of the particle and each super retina unit based on the vertical distance. Sum the similarity weights of each coordinate position information of a super retina unit and the particle to obtain the similarity weight value of the super retina unit. Filter out super retina units whose similarity weight values ​​exceed the set threshold sum_th1. Step 4': Divide each of the selected super retinal units into multiple cell units, calculate the similarity weight value of the cell unit, and select cell units whose similarity weight value exceeds the set threshold sum_th2; determine the linear trajectory of the particle based on the position of the cell unit and output it. The formula for obtaining the similarity weight value of cell unit / super retinal unit based on longitudinal distance is: Where i and j represent the horizontal and vertical serial numbers of the cell unit / super retinal unit, respectively; m and n represent the total number and serial number of the layers of the detector's parallel plane, respectively; The similarity weight value is the coordinate position information of the cell unit / super retina unit in the i-th row and j-th column and the particle detected by the n-th layer detector. This represents the coordinate position information of the particles detected by the nth layer detector and the distance between the cell unit / super retina unit in the i-th row and j-th column along the vertical axis. This is the adjustment factor for the operation; This represents the pixel value of the particle detected by the nth layer detector.

3. The trajectory tracking method according to claim 1 or 2, characterized in that, The formula for obtaining the similarity weight value of cell unit / super retinal unit based on longitudinal distance is: Where i and j represent the horizontal and vertical serial numbers of the cell unit / super retinal unit, respectively; m and n represent the total number and serial number of the layers of the detector's parallel plane, respectively; The similarity weight value is the coordinate position information of the cell in the i-th row and j-th column and the particle detected by the n-th layer detector. This represents the coordinate position information of the particles detected by the nth layer detector and the distance between the cell unit / super retina unit in the i-th row and j-th column along the vertical axis. This is the adjustment factor for the operation; This represents the pixel value of the particle detected by the nth layer detector; d_th is the set distance threshold, and A_th is the set pixel threshold.

4. The trajectory tracking method according to claim 1 or 2, characterized in that, The process of determining and outputting the linear trajectory of the particle based on the position of the cell unit includes: Determine the horizontal and vertical serial numbers i and j of the cell unit / super retina unit, and use the slope and intercept of the cell unit / super retina unit in the i-th row and j-th column as the slope and intercept of the particle linear trajectory.

5. The trajectory tracking method according to claim 2, characterized in that, The trajectory tracking method further includes: Multiple trajectory sample information is obtained by using a detector. The trajectory sample information is the coordinate position information of the pixel points when multiple particles pass through the plane of each layer of detector. Multiple trajectory samples are tracked to find multiple straight-line trajectories, which are then matched with tracks in a standard track library. If only one of the straight-line trajectories matches a track in the standard track library, it is determined to be a perfectly matched track. If at least two of the straight-line trajectories match a track in the standard track library, it is determined to be an over-perfectly matched track. If no track in the standard track library matches any of the straight-line trajectories, it is determined to be a non-matched track. If a straight-line trajectory does not match any track in the standard track library, it is determined to be a false track. The efficiency and purity of the trajectory tracking method are calculated based on the number of four matching trajectories.

6. A particle trajectory tracking device based on the Retina algorithm, characterized in that, The trajectory tracking device includes: a cell array establishment unit, a detector, and a trajectory output unit; The cell array establishment unit is used to divide the two-dimensional space into various spatial units according to a set granularity, and use the spatial units as cell units in the retinal algorithm to establish a cell array. The detector is used to detect and acquire the coordinate position information of the pixel points when a particle passes through the plane of each layer of detectors; The trajectory output unit is used to calculate the similarity weight between the coordinate position information of each particle and each cell unit based on the longitudinal distance, sum the similarity weights of each coordinate position information of a cell unit and the particle to obtain the similarity weight value of the cell unit, traverse all cell units in the cell array, determine the position of the cell unit corresponding to the largest similarity weight value, determine the straight trajectory of the particle based on the position of the cell unit and output it. The formula for obtaining the similarity weight value of cell units based on vertical distance is: Where i and j represent the horizontal and vertical serial numbers of the cell unit, respectively; m and n represent the total number and serial number of the layers in the parallel plane of the detector, respectively. The similarity weight value is the coordinate position information of the cell unit in the i-th row and j-th column and the particle detected by the n-th layer detector. This represents the coordinate position information of the particle detected by the nth layer detector and the distance of the cell unit in the i-th row and j-th column along the vertical axis. This is the adjustment factor for the operation; This represents the pixel value of the particle detected by the nth layer detector.

7. An electronic device, characterized in that, The device includes a memory and a processor, wherein the processor is used to execute computer management programs stored in the memory to implement the steps of the particle trajectory tracking method based on the Retina algorithm as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, It stores a computer management program, which, when executed by a processor, implements the steps of the particle trajectory tracking method based on the Retina algorithm as described in any one of claims 1-5.

Images