A method and apparatus for defining the detection capability of weak targets

Abstract

A method and apparatus for defining the detection capability of weak targets includes: obtaining a detection dataset α1 of a detection system; obtaining a target dataset β1 in a region and using it as an initial benchmark dataset; performing spatiotemporal verification on the target dataset β1 using data from the detection dataset α1; for a preset time interval, when the positional deviation between the target in the detection dataset α1 and the target in the dataset β1 is less than a threshold, determining that the target in the detection dataset α1 is valid, the set of valid targets in the detection dataset α1 is α2, and the set of targets other than valid targets is α3; based on the set α2, verifying the target dataset β1 using a verification method to obtain a dataset β2 containing feature labels; calculating the false alarm rate based on the set α3 and the detection dataset α1; and calculating the detection rate for different energy thresholds based on the label dataset β2 and the set α2.

Description

Technical Field

[0001] This invention relates to a method and apparatus for defining the detection capability of weak targets, belonging to the field of optical remote sensing technology. Background Technology

[0002] Weak target detection technology is widely used in various optical detection systems. Weak target detection is generally defined by the detectivity and false alarm rate under certain energy threshold conditions. The detectivity is defined as the number of points passing the detection threshold divided by the total number of times the target appears; the false alarm rate is the number of false alarms among the points passing the threshold divided by the total number of points.

[0003] In practical applications, detection capability is difficult to define accurately. The main reason is that the energy distribution and statistical results of targets within the field of view are unavailable, making it impossible to determine the corresponding results of the calculated detection rate, false alarm rate, and energy threshold. Traditional methods only perform energy inversion on detected targets, failing to explain whether the energy of undetected targets exceeds the threshold, and thus cannot actually calculate the detection rate under a certain energy threshold. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and solve the problem of defining the detection capability of weak targets.

[0005] The objective of this invention is achieved through the following technical solutions:

[0006] A method for defining the detection capability of weak targets includes:

[0007] Using a certain data source, target detection is performed on targets in a specified area A during the time period T0-T1 to obtain the detection dataset α1 of the detection system;

[0008] Using other data sources, the detection data in the specified region A during the T0-T1 period are statistically analyzed to obtain the target dataset β1 in the region, which is then used as the initial baseline dataset.

[0009] Using the data in the detection dataset α1, spatiotemporal verification is performed on the target dataset β1. For a preset time interval, when the positional deviation between the target in the detection dataset α1 and the target in the dataset β1 is less than a threshold, the target in the detection dataset α1 is determined to be valid. The set of valid targets in the detection dataset α1 is α2, and the set of targets other than valid targets is α3.

[0010] Based on set α2, the target dataset β1 is validated using a validation method to obtain dataset β2 containing feature labels;

[0011] The false alarm rate was calculated based on set α3 and detection dataset α1.

[0012] The detection rate was calculated based on the labeled dataset β2 and the set α2.

[0013] Preferably, the detection dataset α1 contains the target's x-coordinate, y-coordinate, and energy information.

[0014] Preferably, when performing spatiotemporal verification on the target dataset β1, the positional deviation is set according to the positioning capability of the detection system.

[0015] Preferably, the verification method for the target dataset β1 includes: transforming the observed solar vector and the observed vector of the target into the target body coordinate system to obtain the energy label of each target.

[0016] Preferably, the data source includes infrared data source, laser data source, and electromagnetic wave data source.

[0017] Preferably, the target energy is inverted using environmental and observational conditions.

[0018] Preferably, based on the energy threshold, a target set α4 that has exceeded the energy threshold is obtained in β2; the detection rate is calculated using the target set α4 and the set α2.

[0019] A device for defining the detection capability of weak targets, comprising:

[0020] The initial module uses a certain data source to perform target detection on targets in a specified area A during the T0-T1 time period, and obtains the detection dataset α1 of the detection system; using other data sources, it performs statistics on the detection data in the specified area A during the T0-T1 time period, and obtains the target dataset β1 in the area, which is used as the initial baseline dataset.

[0021] The processing module uses the data in the detection dataset α1 to perform spatiotemporal verification on the target dataset β1. For a preset time interval, when the positional deviation between the target in the detection dataset α1 and the target in the dataset β1 is less than a threshold, the target in the detection dataset α1 is determined to be valid. The set of valid targets in the detection dataset α1 is α2, and the set of targets other than valid targets is α3. Based on the set α2, the verification method is used to verify the target dataset β1 to obtain the dataset β2 with feature labels.

[0022] The calculation module calculates the false alarm rate based on set α3 and detection dataset α1; and calculates the detection rate based on label dataset β2 and set α2.

[0023] Compared with the prior art, the present invention has the following advantages:

[0024] (1) This invention breaks through the traditional method of verifying detection rate, false alarm rate and energy inversion separately. Based on the energy inversion model, energy inversion can be achieved for both detected and undetected targets in the field of view, and the detection rate result is calculated more accurately.

[0025] (2) The energy inversion model of the present invention is related to the observation conditions and environmental conditions. Compared with the traditional energy inversion which only considers some environmental conditions, its inversion results are more accurate.

[0026] (3) When performing energy inversion, the present invention can use sample data to correct the characteristic model, thereby further improving the accuracy of the characteristic model and the inversion results of undetected targets.

[0027] (4) This invention utilizes the inversion results to calculate the detection rate under different energy threshold conditions. Combined with the target detection system's own capabilities, the detection threshold can be optimized to achieve improved detection performance. Traditional methods cannot provide the system detection rate under lower detection energy thresholds. Attached Figure Description

[0028] Figure 1 A schematic diagram of the detection information for a specific period of time in the observation area;

[0029] Figure 2 This is a schematic diagram illustrating other information for a specific time period within the observation area. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0031] A method for defining the detection capability of weak targets includes the processes of target identification, model verification, and energy inversion, and its steps are as follows:

[0032] 1) Target detection is performed on the target in the specified area A during the time period T0-T1 to obtain the detection dataset α1 of the detection system.

[0033] 2) Statistical analysis of cooperation information in the specified region A during the T0-T1 time period is performed to obtain the target dataset β1 in the region.

[0034] 3) Perform spatiotemporal verification between the data in dataset α1 and dataset β1. For a specified time interval, if the positional deviation between the target in dataset α1 and the target in dataset β1 is less than a threshold, the target in dataset α1 is considered valid and a model label is obtained. The set of valid targets in the dataset is α2, and the set of targets excluding valid targets is α3.

[0035] 4) Based on α2, the dataset β1 is validated using the validation method to obtain the dataset β2 with feature labels.

[0036] 5) Calculate the false alarm rate using α3 and α1.

[0037] 6) Based on the energy threshold, obtain the target set α4 that exceeds the energy threshold in β2. Calculate the detection rate using α4 and α2. If the number of targets in α4 is m and the number of targets in α2 is n, then the detection rate under this energy threshold condition is n / m.

[0038] The specific method for obtaining the detection dataset α1 in step 1) is as follows: set the detection threshold based on the detection system capability and the observation background; the detection dataset α1 contains the target's x-coordinate, y-coordinate, and energy information.

[0039] In step 3), the positional deviation of the verification method is set to match the positioning capability of the detection system.

[0040] The characteristic verification method in step 4) requires transforming the observed solar vector and observation vector of the target into the target's body coordinate system to obtain the energy label of each target. Accurate inversion of target energy is achieved by utilizing environmental and observational conditions.

[0041] In step 5), the false alarm rate of the target is obtained using measured data and verified data.

[0042] In step 6), the target detection rate is obtained using cooperative data and measured data.

[0043] Data sources include infrared data sources, laser data sources, and electromagnetic wave data sources.

[0044] Example:

[0045] A method for defining the detection capability of weak targets includes:

[0046] In a specific area and at a specific time period, the target trajectory detected by the detection system is as follows: Figure 1 As shown. The target information detected by the detection system includes at least the target's x-coordinate, y-coordinate, and energy at each time point.

[0047] For the area and time period detected by the detection system, the target trajectory obtained based on other information, such as Figure 2 As shown.

[0048] Will Figure 1 The x and y coordinates of the target at the corresponding time and Figure 2 The x and y coordinates of the target are matched, and the matching error at each time point is averaged. If the matching error is less than the matching threshold, the match is considered successful.

[0049] Figure 1 Targets 1, 2, 3 and Figure 2 Targets 6, 7, and 10 were matched, but target 4 could not be matched. Figure 2 Match the target in the middle, then Figure 1Target 4 in the sequence represents a false alarm, and the false alarm rate of this detection can be obtained.

[0050] Based on the elevation angle, azimuth angle, and other environmental conditions such as atmosphere at the time of observation, and according to the radiation characteristic model, Figure 2 Energy calculations are performed on all targets, and the target characteristic model verification process is completed simultaneously.

[0051] By choosing different energy boundaries, we can obtain different energy boundary conditions. Figure 1 and Figure 2 By determining the corresponding targets, the target detection energy boundary of the detection system is obtained, along with the target detectability under that boundary. For example, the energy inversion results for targets 1, 2, and 3 are E1, E2, and E3, while for targets 5, 6, 7, 8, 9, and 10 they are E5, E6, E7, E8, E9, and E10. In the above process, targets 1, 2, and 3 and... Figure 2 Targets 6, 7, and 10 have been matched. If E5, E8, and E9 are all less than the energy threshold, the detection rate is 100%. If one of them is greater than or equal to the energy threshold, the detection rate is 75%. If two of them are greater than or equal to the energy threshold, the detection rate is 60%. If all three are greater than or equal to the energy threshold, the detection rate is 50%.

[0052] A device for defining the detection capability of weak targets, comprising:

[0053] The initial module uses a certain data source to perform target detection on targets in a specified area A during the T0-T1 time period, and obtains the detection dataset α1 of the detection system; using other data sources, it performs statistics on the detection data in the specified area A during the T0-T1 time period, and obtains the target dataset β1 in the area, which is used as the initial baseline dataset.

[0054] The processing module uses the data in the detection dataset α1 to perform spatiotemporal verification on the target dataset β1. For a preset time interval, when the positional deviation between the target in the detection dataset α1 and the target in the dataset β1 is less than a threshold, the target in the detection dataset α1 is determined to be valid. The set of valid targets in the detection dataset α1 is α2, and the set of targets other than valid targets is α3. Based on the set α2, the verification method is used to verify the target dataset β1 to obtain the dataset β2 with feature labels.

[0055] The calculation module calculates the false alarm rate based on set α3 and detection dataset α1; and calculates the detection rate based on label dataset β2 and set α2.

[0056] The contents not described in detail in this specification are common knowledge to those skilled in the art.

[0057] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims

1. A method for defining the detection capability of weak targets, characterized in that, include: Using a certain data source, target detection is performed on targets in a specified area A during the time period T0-T1 to obtain the detection dataset α1 of the detection system; The specific method for obtaining the detection dataset α1 of the detection system is as follows: set the detection threshold based on the detection system's capabilities and the observation background. The detection dataset α1 contains the target's x-coordinate, y-coordinate, and energy information; Using other data sources, the detection data in the specified region A during the T0-T1 period are statistically analyzed to obtain the target dataset β1 in the region, which is then used as the initial baseline dataset. Spatiotemporal verification of the target dataset β1 is performed using data from the detection dataset α1. For a preset time interval, when the positional deviation between the target in the detection dataset α1 and the target in the dataset β1 is less than a threshold, the target in the detection dataset α1 is determined to be valid. The set of valid targets in the detection dataset α1 is α2, and the set of targets other than valid targets is α3. Based on set α2, the target dataset β1 is validated using a validation method to obtain dataset β2 containing feature labels; The verification method for the target dataset β1 includes: transforming the observed solar vector and observation vector of the target to the target body coordinate system to obtain the energy label of each target; and inverting the target energy using environmental and observation conditions. The false alarm rate was calculated based on set α3 and detection dataset α1. Based on the elevation angle, azimuth angle, and atmospheric environmental conditions at the time of observation, energy calculations are performed on all targets in the target dataset β1, and verification is completed simultaneously. Based on the labeled dataset β2 and set α2, the detection rate is calculated; according to the energy threshold, the target set α4 that has exceeded the energy threshold is obtained in β2; using the target set α4 and set α2, the detection rate is calculated.

2. The definition method according to claim 1, characterized in that, When performing spatiotemporal verification on the target dataset β1, the positional deviation is set according to the positioning capability of the detection system.

3. The definition method according to claim 1, characterized in that, Data sources include infrared data sources, laser data sources, and electromagnetic wave data sources.

4. A device for defining the detection capability of weak targets, characterized in that, include: The initial module uses a data source to perform target detection on targets in a specified area A during the time period T0-T1, and obtains the detection dataset α1 of the detection system; use Other data sources are used to statistically analyze the detection data in the specified region A during the T0-T1 time period to obtain the target dataset β1 in the region, which is then used as the initial baseline dataset. The specific method for obtaining the detection dataset α1 of the detection system is as follows: set the detection threshold based on the detection system's capabilities and the observation background. The detection dataset α1 contains the target's x-coordinate, y-coordinate, and energy information; The processing module uses the data in the detection dataset α1 to perform spatiotemporal verification on the target dataset β1; For a preset time interval, when the positional deviation between the target in the detection dataset α1 and the target in the dataset β1 is less than a threshold, the target in the detection dataset α1 is determined to be valid. The set of valid targets in the detection dataset α1 is α2, and the set of targets other than valid targets is α3. Based on set α2, the target dataset β1 is validated using a validation method to obtain dataset β2 containing feature labels; The verification method for the target dataset β1 includes: transforming the observed solar vector and observation vector of the target to the target body coordinate system to obtain the energy label of each target; and inverting the target energy using environmental and observation conditions. The calculation module calculates the false alarm rate based on set α3 and detection dataset α1; it performs energy calculations on all targets in target dataset β1 according to the elevation angle, azimuth angle, and atmospheric environmental conditions at the time of observation, and performs verification at the same time; it calculates the detection rate based on label dataset β2 and set α2; it obtains the target set α4 that has exceeded the energy threshold in β2 according to the energy threshold; and it calculates the detection rate using target set α4 and set α2.

5. The defining device according to claim 4, characterized in that, When performing spatiotemporal verification on the target dataset β1, the positional deviation is set according to the positioning capability of the detection system.

Images