Target flight event prediction method based on infrared dual-band ratio
By performing atmospheric correction and training a multi-channel long short-term memory network model on mid-wave infrared and short-wave infrared time-series data, the problems of missing target flight event prediction methods and the influence of environmental factors were solved, and accurate target flight event prediction and threat identification were achieved.
Patent Information
- Application Number
- CN202511371113.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-24
- Publication Date
- 2026-02-03
AI Technical Summary
Existing technologies lack methods for predicting target flight events, and are greatly affected by external environmental factors, limiting their application scenarios.
By performing atmospheric correction on mid-wave infrared and short-wave infrared time-series data, calculating dual-band radiative ratio data, and using a multi-channel long short-term memory network model for joint training, artificial prior features are extracted to predict target flight events.
It enables accurate prediction of target flight events, reduces the impact of external environmental factors, and supports multi-level defense coordination mechanisms.
Smart Images
Figure CN121456732A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of target event discrimination and prediction technology, specifically to a target flight event prediction method based on infrared dual-band ratio. Background Technology
[0002] During the flight of an aerial target, events such as engine ignition, boost, engine shutdown, and separation may occur. Real-time identification of target events using telemetry data helps detection systems detect threats earlier and quickly initiate mission scheduling and interception procedures, improving the interception system's reaction speed and success rate. Currently, commonly used event identification methods often utilize target flight telemetry image processing for detection. This involves extracting target templates from the images and performing template matching to achieve event detection. However, in real-world scenarios, various influencing factors such as spatial resolution, target-detector distance, detection band, and weather conditions significantly impact target template extraction. Differences in target types and other factors can also affect telemetry-based identification methods.
[0003] In existing technologies, there are many trajectory prediction methods, but there is a lack of methods for target event prediction. Furthermore, the methods for obtaining target event detection data are still based on telemetry image template matching, which is greatly affected by external environmental factors and has limited application scenarios. Summary of the Invention
[0004] The purpose of this invention is to provide a target flight event prediction method based on infrared dual-band ratio, so as to solve the problems of the lack of existing methods for target event prediction, and the fact that target event detection methods are greatly affected by external environmental factors and have limited application scenarios.
[0005] To achieve the above objectives, the present invention is implemented through the following technical solution:
[0006] This invention provides a target flight event prediction method based on infrared dual-band ratio, comprising:
[0007] Step S1: Align the target mid-wave infrared time series data and target short-wave infrared time series data acquired by detectors with different sampling frequencies with the time dimension to obtain mid-wave infrared radiation time series data and short-wave infrared radiation time series data.
[0008] Step S2: By performing atmospheric correction on the mid-wave infrared radiation time series data and the short-wave infrared radiation time series data, the mid-wave body infrared radiation time series data and the short-wave body infrared radiation time series data are obtained, and the ratio of the mid-wave body infrared radiation data to the short-wave body infrared radiation data is calculated to obtain the target dual-band radiation ratio data.
[0009] Step S3: Augment the target dual-band radiation ratio data for different aerial targets and extract artificial prior feature data. Use the augmented target dual-band radiation ratio data and the artificial prior feature data as input data.
[0010] Step S4: Use a multi-channel long short-term memory network model to jointly train the input data to obtain the target flight event prediction results.
[0011] Preferably, in step S1, aligning the target mid-wave infrared time-series data and target short-wave infrared time-series data acquired by detectors with different sampling frequencies along the time dimension specifically includes:
[0012] A linear interpolation method is used to align the mid-wave infrared time series data and the short-wave infrared time series data of the target to the same reference time dimension, thereby obtaining mid-wave infrared radiation time series data and short-wave infrared radiation time series data with the same time interval.
[0013] Preferably, the method of aligning the target mid-wave infrared time-series data and the target short-wave infrared time-series data to the same reference time dimension using linear interpolation specifically includes:
[0014] Step S1.1: For the acquired mid-wave infrared time series data and short-wave infrared time series data of the target, select a set of time series data with more dense time points in the mid-wave infrared time series data and short-wave infrared time series data of the target as a reference sequence;
[0015] Step S1.2: Using a linear interpolation method, the other set of time series data (excluding the reference sequence) from the target mid-wave infrared time series data and the target short-wave infrared time series data is the sequence to be processed. The sequence to be processed is aligned to the reference sequence.
[0016] The sequence to be processed is (t1,L1), (t2,L2),...,(t... n ,L n ), where t1 < t2 < ... < t n If a certain time point t on the reference sequence is located at time point t on the sequence to be processed... i and t i+1 Between, i.e., t i ≤t≤t i+1 The radiation value L(t) at time point t is calculated using the following formula:
[0017]
[0018] Among them, L i+1 Represents time point t on the sequence to be processed i+1 Corresponding radiation value; Li Represents time point t on the sequence to be processed i The corresponding radiation value;
[0019] The radiation values corresponding to each time point on the reference sequence are calculated sequentially, and the obtained (t, L(t)) is inserted into the sequence to be processed to obtain the aligned sequence to be processed.
[0020] Preferably, in step S2, atmospheric correction is performed on the mid-wave infrared radiation time-series data and the short-wave infrared radiation time-series data to obtain the mid-wave body infrared radiation time-series data and the short-wave body infrared radiation time-series data, and the ratio of the mid-wave body infrared radiation data to the short-wave body infrared radiation data is calculated to obtain the target dual-band radiation ratio data, specifically including:
[0021] Establish a radiative transfer model and calculate atmospheric transmittance and atmospheric radiance parameters;
[0022] Based on the atmospheric transmittance and atmospheric radiance parameters, the atmospheric-corrected mid-wave infrared radiation time series data of the target body are calculated using the radiative transfer equation. and shortwave infrared radiation time series data
[0023] The ratio of mid-wave infrared radiation data to short-wave infrared radiation data of the target body was calculated.
[0024] Preferably, the augmentation of the target dual-band radiation ratio data for different aerial targets in step S3 specifically includes:
[0025] The target dual-band radiometric ratio data is augmented using one or a combination of time-slice sliding window, noise addition, cropping, and stitching.
[0026] Preferably, step S3, which involves extracting artificial prior feature data, specifically includes:
[0027] Extract the statistical and morphological characteristics of the target's dual-band radiation ratio under the event.
[0028] Preferably, the statistical characteristics include: mean, variance, maximum value, and minimum value.
[0029] Preferably, the morphological features include kurtosis, skewness, and curvature.
[0030] Preferably, in step S4, a multi-channel long short-term memory network model is used to jointly train the input data to obtain the target flight event prediction result, which specifically includes:
[0031] The input data is reshaped to construct standardized input data;
[0032] A multi-channel Long Short-Term Memory (LSTM) network model is constructed. The event information of the target at the current moment is used as the label, and the standardized input data of the target is used as the input. The data is fused and spliced through the output layer of the multi-channel LSTM network model, and the predicted data is output through the fully connected layer of the multi-channel LSTM network model.
[0033] Based on the event information of the target at the current moment and the prediction data, the current event state of the target is predicted to obtain the target flight event prediction result.
[0034] Preferably, the multi-channel long short-term memory network model is a dual-channel long short-term memory network model. The augmented target dual-band radiation ratio data and the artificial prior feature data are respectively input into the dual-channel long short-term memory network model, and the dual-channel long short-term memory network features are output after passing through the output layer of the multi-channel long short-term memory network model.
[0035] Compared with the prior art, the present invention has the following beneficial effects:
[0036] This invention addresses the challenge of predicting target flight events by incorporating the infrared dual-band ratio feature information of targets into event judgment and prediction. It also augments limited small-sample time-series data and outputs the dual-band ratio information and artificial features together into a multi-channel neural network. This helps to detect abnormal flight target activities in advance, thereby enabling the identification and prediction of non-cooperative target threats and supporting multi-level defense collaboration mechanisms. Attached Figure Description
[0037] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description will be briefly introduced below. Obviously, the drawings described below are one embodiment of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort:
[0038] Figure 1 A flowchart of a target flight event prediction method based on infrared dual-band ratio provided in an embodiment of the present invention;
[0039] Figure 2 This is a flowchart of dual-band data time dimension alignment provided in an embodiment of the present invention;
[0040] Figure 3 This is a diagram illustrating the dual-band ratio acquisition and processing provided in an embodiment of the present invention.
[0041] Figure 4A block diagram illustrating the preparation of dual-band ratio input data according to an embodiment of the present invention;
[0042] Figure 5 This is a flowchart of flight event prediction provided in an embodiment of the present invention. Detailed Implementation
[0043] The following is in conjunction with the appendix Figure 1-5 The following detailed description further illustrates the target flight event prediction method based on infrared dual-band ratio proposed in this invention. The advantages and features of this invention will become clearer from the following description. It should be noted that the accompanying drawings are in a very simplified form and use non-precise proportions, used only to facilitate and clearly illustrate the purpose of the embodiments of this invention. Please refer to the accompanying drawings to make the objectives, features, and advantages of this invention more apparent and understandable. It should be understood that the structures, proportions, sizes, etc., depicted in the accompanying drawings are only used to complement the content disclosed in the specification, for those skilled in the art to understand and read, and are not intended to limit the implementation conditions of this invention. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportional relationships, or adjustments to the size, without affecting the effects and objectives achieved by this invention, should still fall within the scope of the technical content disclosed in this invention.
[0044] refer to Figure 1 As shown, this embodiment provides a target flight event prediction method based on infrared dual-band ratio, including the following steps:
[0045] Step S1: Align the target mid-wave infrared time series data and target short-wave infrared time series data acquired by detectors with different sampling frequencies with the time dimension to obtain mid-wave infrared radiation time series data and short-wave infrared radiation time series data.
[0046] refer to Figure 2 As shown, a linear interpolation method is used to align the target's mid-wave infrared time-series data and the target's short-wave infrared time-series data to the same reference time dimension, obtaining mid-wave infrared radiation time-series data and short-wave infrared radiation time-series data with the same time interval. Specifically, this includes:
[0047] Step S1.1: For the acquired mid-wave infrared time series data and short-wave infrared time series data of the target, select a set of time series data with more dense time points in the mid-wave infrared time series data and short-wave infrared time series data of the target as a reference sequence.
[0048] Step S1.2: Using a linear interpolation method, the target mid-wave infrared time series data and the target short-wave infrared time series data, excluding the reference sequence, are another set of time series data to be processed. The sequence to be processed is aligned to the reference sequence.
[0049] The sequence to be processed is (t1,L1), (t2,L2),...,(t... n ,L n ), where t1 < t2 < ... < t n If a certain time point t on the reference sequence is located at time point t on the sequence to be processed... i and t i+1 Between, i.e., t i ≤t≤t i+1 The radiation value L(t) at time point t is calculated using the following formula:
[0050]
[0051] Among them, L i+1 Represents time point t on the sequence to be processed i+1 Corresponding radiation value; L i Represents time point t on the sequence to be processed i The corresponding radiation value.
[0052] The radiation values corresponding to each time point in the reference sequence are calculated sequentially, and the obtained (t, L(t)) is inserted into the sequence to be processed to obtain the aligned sequence to be processed, thereby obtaining the target mid-wave infrared radiation time series data of the same time interval. With shortwave infrared radiation time series data
[0053] Step S2: Reference Figure 3 As shown, by performing atmospheric correction on the mid-wave infrared radiation time-series data and the short-wave infrared radiation time-series data, the mid-wave body infrared radiation time-series data and the short-wave body infrared radiation time-series data are obtained, and the ratio of the mid-wave body infrared radiation data to the short-wave body infrared radiation data is calculated to obtain the target dual-band radiation ratio data, which specifically includes:
[0054] A radiative transfer model was established using atmospheric radiative transfer simulation software to calculate atmospheric transmittance and atmospheric radiance parameters. Based on the atmospheric transmittance and atmospheric radiance parameters, atmospherically corrected mid-wave infrared radiation time-series data of the target body were calculated using the radiative transfer equation. and shortwave infrared radiation time series data The ratio of mid-wave infrared radiation data to short-wave infrared radiation data of the target body was calculated.
[0055] Step S3: Reference Figure 4 As shown, the target dual-band radiation ratio data for different aerial targets is augmented, and artificial prior feature data is extracted. Both the augmented target dual-band radiation ratio data and the artificial prior feature data are used as input data.
[0056] The augmentation of the target dual-band radiometric ratio data for different aerial targets specifically includes: augmenting limited small-sample time-series data by using one or any combination of time-slice sliding window, adding noise, cropping, and splicing to augment the target dual-band radiometric ratio data.
[0057] The time-slice sliding window method first sets the window size (e.g., 50 time points) and the sliding step size (e.g., 20 time points). The sliding window is then used to slide across the time series to extract each time segment as a new time series sample. Adding noise can add random noise to the value of each time point in the time series. In this example, a small Gaussian noise is added. A Gaussian noise with a mean of 0 and a standard deviation of 0.01 is randomly added to the value of each time point to simulate measurement error or environmental changes.
[0058] Step S3 involves extracting prior artificial feature data, specifically including: extracting statistical and morphological features of the target's dual-band radiation ratio under the event. The statistical features include: mean, variance, maximum, and minimum values. The morphological features include: kurtosis, skewness, and curvature.
[0059] Step S4: Reference Figure 5 As shown, a multi-channel long short-term memory network model is used to jointly train the input data to obtain the target flight event prediction result, which specifically includes:
[0060] The input data is reshaped to construct standardized input data; a multi-channel long short-term memory network model is constructed, using the event information of the target at the current moment as a label and the standardized input data of the target as input. The data is fused and spliced through the output layer of the multi-channel long short-term memory network model, and the predicted data is output through the fully connected layer of the multi-channel long short-term memory network model; based on the event information of the target at the current moment and the predicted data, the current event state of the target is predicted to obtain the target flight event prediction result.
[0061] In this embodiment, the multi-channel long short-term memory (LSTM) network model is a dual-channel LSTM network model. The augmented target dual-band radiometric ratio data and the artificial prior feature data are input into the dual-channel LSTM network model. After passing through the input gate, forget gate, and hidden layers of the dual-channel LSTM network model, and then through the output layer of the multi-channel LSTM network model, dual-channel LSTM network features are output. The dual-channel LSTM network features output from the output layer are fused and concatenated, and then, through a fully connected layer and a normalized exponential function, the current event state is predicted based on the input historical event state information.
[0062] This embodiment provides a target flight event prediction method based on infrared dual-band ratio. Due to differences in atmospheric absorption and transmission characteristics across different bands, the target infrared dual-band ratio varies significantly under different flight events. Utilizing a multi-channel long short-term memory (LSTM) network model, the potential patterns of flight events changing with infrared dual-band ratio data can be extracted more effectively. Therefore, this embodiment collaboratively applies dual-band ratio data with distinct characteristics and extracted artificial features to address the challenge of target flight event prediction. By incorporating target infrared dual-band ratio feature information into event judgment and prediction, and performing data augmentation on limited small-sample time-series data, the dual-band ratio information and artificial features are jointly output to a multi-channel neural network. This helps to detect abnormal flight target activity in advance, thereby enabling the identification and prediction of non-cooperative target threats and supporting multi-layered defense coordination mechanisms.
[0063] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0064] It should be noted that the apparatus and methods disclosed in the embodiments herein can also be implemented in other ways. The apparatus embodiments described above are merely illustrative; for example, the flowcharts and block diagrams in the accompanying drawings show the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments herein. In this regard, each block in a flowchart or block diagram may represent a module, program, or part of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system to perform the specified function or action, or can be implemented using a combination of dedicated hardware and computer instructions.
[0065] In addition, the functional modules in the various embodiments of this article can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.
[0066] Although the present invention has been described in detail through the preferred embodiments above, it should be understood that the above description should not be considered as a limitation of the present invention. Various modifications and substitutions to the present invention will be apparent to those skilled in the art after reading the above description. Therefore, the scope of protection of the present invention should be defined by the appended claims.
Claims
1. A target flight event prediction method based on infrared dual-band ratio, characterized in that, include: Step S1: Align the target mid-wave infrared time series data and target short-wave infrared time series data acquired by detectors with different sampling frequencies with the time dimension to obtain mid-wave infrared radiation time series data and short-wave infrared radiation time series data. Step S2: By performing atmospheric correction on the mid-wave infrared radiation time series data and the short-wave infrared radiation time series data, the mid-wave body infrared radiation time series data and the short-wave body infrared radiation time series data are obtained, and the ratio of the mid-wave body infrared radiation data to the short-wave body infrared radiation data is calculated to obtain the target dual-band radiation ratio data. Step S3: Augment the target dual-band radiation ratio data for different aerial targets and extract artificial prior feature data. Use the augmented target dual-band radiation ratio data and the artificial prior feature data as input data. Step S4: Use a multi-channel long short-term memory network model to jointly train the input data to obtain the target flight event prediction results.
2. The target flight event prediction method based on infrared dual-band ratio as described in claim 1, characterized in that, In step S1, aligning the target mid-wave infrared time-series data and target short-wave infrared time-series data acquired by detectors with different sampling frequencies along the time dimension specifically includes: A linear interpolation method is used to align the mid-wave infrared time series data and the short-wave infrared time series data of the target to the same reference time dimension, thereby obtaining mid-wave infrared radiation time series data and short-wave infrared radiation time series data with the same time interval.
3. The target flight event prediction method based on infrared dual-band ratio as described in claim 2, characterized in that, The method of aligning the target's mid-wave infrared time-series data and short-wave infrared time-series data to the same reference time dimension using linear interpolation specifically includes: Step S1.1: For the acquired mid-wave infrared time series data and short-wave infrared time series data of the target, select a set of time series data with more dense time points in the mid-wave infrared time series data and short-wave infrared time series data of the target as a reference sequence; Step S1.2: Using a linear interpolation method, the other set of time series data (excluding the reference sequence) from the target mid-wave infrared time series data and the target short-wave infrared time series data is the sequence to be processed. The sequence to be processed is aligned to the reference sequence. The sequence to be processed is (t1,L1), (t2,L2),...,(t... n ,L n ), where t1 < t2 < ... < t n If a certain time point t on the reference sequence is located at time point t on the sequence to be processed... i and t i+1 Between, i.e., t i ≤t≤t i+1 The radiation value L(t) at time point t is calculated using the following formula: Among them, L i+1 Represents time point t on the sequence to be processed i+1 Corresponding radiation value; L i Represents time point t on the sequence to be processed i The corresponding radiation value; The radiation values corresponding to each time point on the reference sequence are calculated sequentially, and the obtained (t, L(t)) is inserted into the sequence to be processed to obtain the aligned sequence to be processed.
4. The target flight event prediction method based on infrared dual-band ratio as described in claim 3, characterized in that, In step S2, atmospheric correction is performed on the mid-wave infrared radiation time-series data and the short-wave infrared radiation time-series data to obtain the mid-wave body infrared radiation time-series data and the short-wave body infrared radiation time-series data. The ratio of the mid-wave body infrared radiation data to the short-wave body infrared radiation data is then calculated to obtain the target dual-band radiation ratio data. Specifically, this includes: Establish a radiative transfer model and calculate atmospheric transmittance and atmospheric radiance parameters; Based on the atmospheric transmittance and atmospheric radiance parameters, the atmospheric-corrected mid-wave infrared radiation time series data of the target body are calculated using the radiative transfer equation. and shortwave infrared radiation time series data The ratio of mid-wave infrared radiation data to short-wave infrared radiation data of the target body was calculated.
5. The target flight event prediction method based on infrared dual-band ratio as described in claim 1, characterized in that, Step S3 involves augmenting the target dual-band radiation ratio data for different aerial targets, specifically including: The target dual-band radiometric ratio data is augmented using one or a combination of time-slice sliding window, noise addition, cropping, and stitching.
6. The target flight event prediction method based on infrared dual-band ratio as described in claim 1, characterized in that, The step S3 involves extracting prior artificial feature data, specifically including: Extract the statistical and morphological characteristics of the target's dual-band radiation ratio under the event.
7. The target flight event prediction method based on infrared dual-band ratio as described in claim 6, characterized in that, The statistical characteristics include: mean, variance, maximum value, and minimum value.
8. The target flight event prediction method based on infrared dual-band ratio as described in claim 6, characterized in that, The morphological features include kurtosis, skewness, and curvature.
9. The target flight event prediction method based on infrared dual-band ratio as described in claim 1, characterized in that, Step S4 employs a multi-channel long short-term memory network model to jointly train the input data, obtaining the target flight event prediction result, which specifically includes: The input data is reshaped to construct standardized input data; A multi-channel long short-term memory network model is constructed, with the event information of the target at the current moment as the label and the standardized input data of the target as the input. The data is fused and spliced through the output layer of the multi-channel long short-term memory network model, and the predicted data is output through the fully connected layer of the multi-channel long short-term memory network model. Based on the event information of the target at the current moment and the prediction data, the current event state of the target is predicted to obtain the target flight event prediction result.
10. The target flight event prediction method based on infrared dual-band ratio as described in claim 9, characterized in that, The multi-channel long short-term memory network model is a dual-channel long short-term memory network model. The augmented target dual-band radiation ratio data and the artificial prior feature data are respectively input into the dual-channel long short-term memory network model. After passing through the output layer of the multi-channel long short-term memory network model, the dual-channel long short-term memory network features are output.