A method and apparatus for determining moving targets in infrared images

Abstract

This invention discloses a method and apparatus for determining moving targets in infrared images. The method for determining moving targets in infrared images includes: acquiring all detected targets in the current frame image, defining each detected target as a baseline target, and calculating the relative velocities of other detected targets relative to the baseline target; if the relative velocities of all other detected targets relative to the baseline target are greater than a first velocity threshold, then voting for the baseline target once; if there is one and only one other detected target whose relative velocity relative to the baseline target is greater than a second velocity threshold, then voting for that other detected target once; statistically analyzing the votes for each detected target, and determining the detected target with the most votes as the moving target, while the other detected targets are false alarms. By using this invention, calculating the relative velocities between detected targets, and performing a traversal voting process according to preset voting rules, the detected target with the most votes is determined to be the moving target, thus improving the accuracy of moving target recognition.

Description

Technical Field

[0001] This invention relates to the field of target recognition technology, and in particular to a method and apparatus for determining moving targets in infrared images. Background Technology

[0002] Infrared target detection technology primarily utilizes the difference in infrared radiation between the background and the target for target detection. A crucial performance indicator for infrared long-range early warning and detection systems is their ability to detect small targets, with moving targets being the primary focus of these systems. When the target is far from the detector, its spectral energy is transmitted through the atmosphere. Due to atmospheric disturbances, optical scattering, and diffraction, the spectral irradiance received by the detector target surface is very low, resulting in a low signal-to-noise ratio. Furthermore, in downward-looking scenarios, ground conditions are often complex, with numerous bright spots on the ground that closely resemble the target. Additionally, distant targets have a small area in infrared images, making them difficult to distinguish from blind and flash elements, leading to numerous false alarms caused by similar targets and hindering accurate identification of moving targets. Summary of the Invention

[0003] This invention provides a method and apparatus for determining moving targets in infrared images, which at least solves the problem of inaccurate identification of moving targets in the prior art.

[0004] The first aspect of this invention provides a method for determining moving targets in infrared images, comprising:

[0005] All detected targets in the current frame image are obtained, and each detected target is defined as a reference target in turn. The relative velocities of other detected targets relative to the reference targets are calculated.

[0006] If the relative speeds of the other detected targets relative to the benchmark target are all greater than the first speed threshold, then the benchmark target is voted once.

[0007] If there is one and only one other detected target with a relative velocity greater than the reference target, then vote once for that other detected target;

[0008] The votes for each of the aforementioned detection targets are tallied, and the target with the most votes is determined to be the moving target, while the other detection targets are considered false alarms.

[0009] According to some embodiments of the present invention, calculating the relative velocity of other detected targets relative to the reference target includes:

[0010] Obtain the historical coordinate values ​​of the detected target in each historical frame image;

[0011] Based on the historical coordinate values, the relative velocities of other detected targets relative to the baseline target are calculated.

[0012] According to some embodiments of the present invention, calculating the relative velocity of other detected targets relative to the reference target based on the historical coordinate values ​​includes:

[0013] The relative velocities V of other detected targets relative to the reference target are calculated based on Formula 1:

[0014] V = |(S2-S1) / t| (Formula 1)

[0015] Wherein, S1 represents the distance between the other detected target and the reference target when they were first located in the same historical frame image, S2 represents the distance between the other detected target and the reference target in the current frame image, the distance unit of S1 and S2 is pixels, and t represents the difference in the number of frames between the historical frame image and the current frame image.

[0016] According to some embodiments of the present invention, calculating the relative velocity of other detected targets relative to the reference target includes:

[0017] Calculate the distance between other detected targets and the reference target, and calculate the relative velocity of the other detected target relative to the reference target if and only if the distance is less than or equal to a preset distance.

[0018] According to some embodiments of the present invention, obtaining the historical coordinate values ​​of the detected target in each historical frame image includes:

[0019] Obtain a second threshold of historical frame images whose distance from the current frame image is greater than a first threshold of historical frame images;

[0020] Obtain the historical coordinate values ​​of the detected target in each of the second threshold historical frame images.

[0021] According to some embodiments of the present invention, the method further includes:

[0022] The number of consecutive frames in which the detected target is identified as a moving target is determined. If and only if the number of consecutive frames is greater than a frame count threshold, the moving target is identified as a real moving target.

[0023] According to some embodiments of the present invention, the first speed threshold and the second speed threshold are equal, both being 0.2 pixels / frame.

[0024] A second aspect of the present invention provides a device for determining moving targets in infrared images, comprising:

[0025] The motion information extraction unit is used to acquire all detected targets in the current frame image, define each detected target as a reference target in sequence, and calculate the relative velocity of other detected targets relative to the reference target.

[0026] The voting statistics unit is used to vote according to the following rules:

[0027] If the relative speeds of the other detected targets relative to the benchmark target are all greater than the first speed threshold, then the benchmark target is voted once.

[0028] If there is one and only one other detected target with a relative velocity greater than the reference target, then vote once for that other detected target;

[0029] The voting statistics unit is also used to count the votes for each of the detected targets after the voting is completed, and to determine that the detected target with the most votes is a moving target, while the other detected targets are false alarms.

[0030] A third aspect of the present invention provides a device for determining moving targets in infrared images, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of the method for determining moving targets in infrared images as described in any one of the first aspects of the present invention.

[0031] A fourth aspect of the present invention provides a computer-readable storage medium storing an information transmission implementation program, which, when executed by a processor, implements the steps of the method for determining moving targets in an infrared image as described in any one of the first aspects of the embodiment.

[0032] The technical solution of this invention calculates the relative speed between detected targets and performs a traversal voting process according to preset voting rules, identifying the target with the most votes as the moving target. This significantly reduces the false alarm rate in the detection of weak moving targets and improves the accuracy of moving target identification.

[0033] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0034] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of the embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. In the drawings:

[0035] Figure 1 This is a flowchart illustrating the method for determining moving targets in infrared images in an embodiment of the present invention;

[0036] Figure 2 This is a schematic diagram of the relative speed corresponding to a voting situation in an embodiment of the present invention;

[0037] Figure 3 This is a schematic diagram of the relative speed corresponding to another voting scenario in an embodiment of the present invention. Detailed Implementation

[0038] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0039] The first aspect of this invention provides a method for determining moving targets in infrared images, comprising:

[0040] All detected targets in the current frame image are acquired, and each detected target is defined as a baseline target. The relative velocities of other detected targets relative to the baseline targets are calculated. For ease of statistical analysis, a relative velocity statistics table can be created for each baseline target.

[0041] For example, if three targets are detected in the current frame image, namely the first target, the second target, and the third target, the relative velocities of the first, second, and third targets with respect to the first target are calculated first, and a first relative velocity statistics table is created. Then, using the second target as the reference target, the relative velocities of the first, third, and second targets with respect to the second target are calculated, and a second relative velocity statistics table is created. Finally, using the third target as the reference target, the relative velocities of the first, second, and third targets with respect to the third target are calculated, and a third relative velocity statistics table is created.

[0042] According to the voting rules described below, a vote is taken for the detection target based on each relative velocity statistics table.

[0043] If the relative velocities of other detected targets relative to the baseline target are all greater than the first velocity threshold, then a vote is cast for this baseline target. For example, Figure 2 The table shown presents the relative velocity statistics for Bbox1 as the baseline target, with a first velocity threshold of 0.2 pixels / frame. (Reference) Figure 2Bbox1 is the baseline target, and Bbox2, Bbox3, Bbox4, and Bbox5 are other detection targets. If the relative speed of Bbox2, Bbox3, Bbox4, and Bbox5 with Bbox1 is greater than 0.2 pixels / frame, then Bbox1 is voted once as the baseline target.

[0044] If there is one and only one other detected target with a relative velocity greater than the second velocity threshold relative to the reference target, then vote once for that other detected target. For example, referencing... Figure 3 , Figure 3 The table shows the relative velocity statistics for Bbox1 as the baseline target, with a second velocity threshold of 0.2 pixels / frame. Bbox1 is the baseline target, and Bbox2, Bbox3, Bbox4, and Bbox5 are the other detection targets. If only Bbox2's relative velocity is greater than 0.2 pixels / frame, then Bbox2 receives one vote from the other detection targets.

[0045] The system iterates through all relative velocity statistics tables to conduct voting, calculates the votes for each detection target, and determines the detection target with the most votes as the moving target, while the other detection targets are considered false alarms.

[0046] The technical solution of this invention calculates the relative speed between detected targets and performs a traversal voting process according to preset voting rules, identifying the target with the most votes as a moving target. This significantly reduces the false alarm rate in detecting weak moving targets and improves the accuracy of moving target recognition.

[0047] Based on the above embodiments, further variant embodiments are proposed. It should be noted that, in order to keep the description brief, only the differences from the above embodiments are described in each variant embodiment.

[0048] According to some embodiments of the present invention, calculating the relative velocity of other detected targets relative to the reference target includes:

[0049] Obtain the historical coordinate values ​​of the detected targets in each historical frame image. For example, a batch can be created for each detected target, including all historical coordinate positions of the detected target and the corresponding time frame number, which facilitates subsequent matching and calculation of relative speed between targets.

[0050] Based on the historical coordinate values, the relative velocities of other detected targets relative to the baseline target are calculated.

[0051] According to some embodiments of the present invention, calculating the relative velocity of other detected targets relative to the reference target based on the historical coordinate values ​​includes:

[0052] The relative velocities V of other detected targets relative to the reference target are calculated based on Formula 1:

[0053] V = |(S2-S1) / t| (Formula 1)

[0054] Wherein, S1 represents the distance between the other detected target and the reference target when they were first located in the same historical frame image, S2 represents the distance between the other detected target in S1 and the reference target in the current frame image, the distance between S1 and S2 is in pixels, and t represents the difference in the number of frames between the historical frame image and the current frame image.

[0055] According to some embodiments of the present invention, calculating the relative velocity of other detected targets relative to the reference target includes: calculating the distance between the other detected target and the reference target, and calculating the relative velocity of the other detected target relative to the reference target if and only if the distance is less than or equal to a preset distance. For example, if the distance between two detected targets is less than or equal to 20 pixel values ​​of the current frame image, the two detected targets are identified as associative targets, and their relative velocity is then calculated.

[0056] According to some embodiments of the present invention, for detection targets that cannot be associated, a separate batch is set up for each target, and the location information and corresponding time frame number of the detection target are stored in the batch so that the relative velocity can be calculated when association can be generated later.

[0057] According to some embodiments of the present invention, after a moving target is determined, it is reported as an output to the control center or terminal. When the moving target changes, the reporting record is updated.

[0058] According to some embodiments of the present invention, obtaining the historical coordinate values ​​of the detected targets in each historical frame image includes: obtaining a second threshold of historical frame images whose distance from the current frame image is greater than a first threshold of historical frame images. The historical coordinate values ​​of the detected targets in each of the second threshold of historical frame images are then obtained. For example, when calculating the relative velocity between detected targets, frames whose distance from the current frame is greater than half the total number of stored frames are selected for calculating the relative velocity. For example, if the current frame is the nth frame, and the information from the previous 100 frames is stored, then when calculating motion information, information from frames n-100 to n-50 is used for calculation. This can extract the positional differences inherent in the reported results of each frame, eliminating errors caused by factors such as camera shake.

[0059] According to some embodiments of the present invention, the method further includes counting the number of consecutive frames in which the detected target is determined to be a moving target, and determining it as a real moving target if and only if the number of consecutive frames is greater than a frame number threshold; otherwise, determining it as a false alarm.

[0060] According to some embodiments of the present invention, the frame number threshold is 100.

[0061] According to some embodiments of the present invention, the frame number threshold should be greater than twice the number of frames for saving voting results, wherein the number of frames for saving voting results is the number of frames used when calculating voting results. This can prevent erroneous voting results within a few frames from continuously affecting the statistical results over a long period, and effectively reduce the probability of reporting false alarm targets.

[0062] According to some embodiments of the present invention, the first speed threshold and the second speed threshold are equal, both being 0.2 pixels / frame.

[0063] A second aspect of the present invention provides a device for determining moving targets in infrared images, comprising:

[0064] The motion information extraction unit is used to acquire all detected targets in the current frame image, define each detected target as a reference target in sequence, and calculate the relative velocity of other detected targets relative to the reference target.

[0065] For example, if three targets are detected in the current frame image, namely the first target, the second target, and the third target, the relative velocities of the first, second, and third targets with respect to the first target are calculated first, and a first relative velocity statistics table is created. Then, using the second target as the reference target, the relative velocities of the first, third, and second targets with respect to the second target are calculated, and a second relative velocity statistics table is created. Finally, using the third target as the reference target, the relative velocities of the first, second, and third targets with respect to the third target are calculated, and a third relative velocity statistics table is created.

[0066] The voting statistics unit is used to vote according to the following rules:

[0067] If the relative velocities of the other detected targets relative to the benchmark target are all greater than a first velocity threshold, then the benchmark target receives one vote. For example, Figure 2 The table shown presents the relative velocity statistics for Bbox1 as the baseline target, with a first velocity threshold of 0.2 pixels / frame. (Reference) Figure 2 Bbox1 is the baseline target, and Bbox2, Bbox3, Bbox4, and Bbox5 are other detection targets. If the relative speed of Bbox2, Bbox3, Bbox4, and Bbox5 with Bbox1 is greater than 0.2 pixels / frame, then Bbox1 is voted once as the baseline target.

[0068] If there is one and only one other detected target with a relative velocity greater than the reference target, then that other detected target is voted on once. For example, refer to Figure 3 , Figure 3The table shows the relative velocity statistics for Bbox1 as the baseline target, with a second velocity threshold of 0.2 pixels / frame. Bbox1 is the baseline target, and Bbox2, Bbox3, Bbox4, and Bbox5 are the other detection targets. If only Bbox2's relative velocity is greater than 0.2 pixels / frame, then Bbox2 receives one vote from the other detection targets.

[0069] The voting statistics unit is also used to count the votes for each of the detected targets after the voting is completed, and to determine that the detected target with the most votes is a moving target, while the other detected targets are false alarms.

[0070] The technical solution of this invention calculates the relative speed between detected targets and performs a traversal voting process according to preset voting rules, identifying the target with the most votes as a moving target. This significantly reduces the false alarm rate in detecting weak moving targets and improves the accuracy of moving target recognition.

[0071] Based on the above embodiments, further variant embodiments are proposed. It should be noted that, in order to keep the description brief, only the differences from the above embodiments are described in each variant embodiment.

[0072] According to some embodiments of the present invention, calculating the relative velocity of other detected targets relative to the reference target includes:

[0073] Obtain the historical coordinate values ​​of the detected targets in each historical frame image. For example, a batch can be created for each detected target, including all historical coordinate positions of the detected target and the corresponding time frame number, which facilitates subsequent matching and calculation of relative speed between targets.

[0074] Based on the historical coordinate values, the relative velocities of other detected targets relative to the baseline target are calculated.

[0075] According to some embodiments of the present invention, calculating the relative velocity of other detected targets relative to the reference target based on the historical coordinate values ​​includes:

[0076] The relative velocities V of other detected targets relative to the reference target are calculated based on Formula 1:

[0077] V = |(S2-S1) / t| (Formula 1)

[0078] Wherein, S1 represents the distance between the other detected target and the reference target when they were first located in the same historical frame image, S2 represents the distance between the other detected target in S1 and the reference target in the current frame image, the distance between S1 and S2 is in pixels, and t represents the difference in the number of frames between the historical frame image and the current frame image.

[0079] According to some embodiments of the present invention, calculating the relative velocity of other detected targets relative to the reference target includes: calculating the distance between the other detected target and the reference target, and calculating the relative velocity of the other detected target relative to the reference target if and only if the distance is less than or equal to a preset distance. For example, if the distance between two detected targets is less than or equal to 20 pixel values ​​of the current frame image, the two detected targets are identified as associative targets, and their relative velocity is then calculated.

[0080] According to some embodiments of the present invention, for detection targets that cannot be associated, a separate batch is set up for each target, and the location information and corresponding time frame number of the detection target are stored in the batch so that the relative velocity can be calculated when association can be generated later.

[0081] According to some embodiments of the present invention, after a moving target is determined, it is reported as an output to the control center or terminal. When the moving target changes, the reporting record is updated.

[0082] According to some embodiments of the present invention, obtaining the historical coordinate values ​​of the detected targets in each historical frame image includes: obtaining a second threshold of historical frame images whose distance from the current frame image is greater than a first threshold of historical frame images. The historical coordinate values ​​of the detected targets in each of the second threshold of historical frame images are then obtained. For example, when calculating the relative velocity between detected targets, frames whose distance from the current frame is greater than half the total number of stored frames are selected for calculating the relative velocity. For example, if the current frame is the nth frame, and the information from the previous 100 frames is stored, then when calculating motion information, information from frames n-100 to n-50 is used for calculation. This can extract the positional differences inherent in the reported results of each frame, eliminating errors caused by factors such as camera shake.

[0083] According to some embodiments of the present invention, the voting statistics unit is further configured to count the number of consecutive frames in which the detected target is determined to be a moving target, and determine it as a real moving target if and only if the number of consecutive frames is greater than a frame number threshold; otherwise, determine it as a false alarm.

[0084] According to some embodiments of the present invention, the frame number threshold is 100.

[0085] According to some embodiments of the present invention, the frame number threshold should be greater than twice the number of frames for saving voting results, wherein the number of frames for saving voting results is the number of frames used when calculating voting results. This can prevent erroneous voting results within a few frames from continuously affecting the statistical results over a long period, and effectively reduce the probability of reporting false alarm targets.

[0086] According to some embodiments of the present invention, the first speed threshold and the second speed threshold are equal, both being 0.2 pixels / frame.

[0087] The following describes in detail a method for determining moving targets in infrared images using a specific embodiment.

[0088] It is understood that the following description is merely illustrative and not a specific limitation of the invention. Any similar structures or variations thereof employing the present invention should be included within the scope of protection of the present invention.

[0089] In this embodiment, reference Figure 1 There are N detected targets reported in the current frame. First, a corresponding batch is established for each detected target. Different batches store the position and time frame number of the corresponding detected target in different frame images. The detected targets in the current frame are associated, and the relevant information of the associatable detected targets is extracted from their corresponding batches. Each detected target is defined as a baseline target, and the relative velocity V of other detected targets relative to the baseline target is calculated according to Formula 1 as follows:

[0090] V = |(S2-S1) / t| (Formula 1)

[0091] Wherein, S1 represents the distance between the other detected target and the reference target when they were first located in the same historical frame image, S2 represents the distance between the other detected target in S1 and the reference target in the current frame image, the distance unit of S1 and S2 is pixels, and t represents the frame number difference between the historical frame image and the current frame image.

[0092] After calculating the relative velocity, a relative velocity statistics table is created for each benchmark target, for reference. Figure 2 or Figure 3 Iterate through the relative velocity statistics table and vote according to the following rules:

[0093] If the relative speed of the other detected targets relative to the benchmark target is greater than the first speed threshold, then the benchmark target is voted once.

[0094] If there is one and only one other detection target whose relative velocity to the reference target is greater than the second velocity threshold, then vote once for that other detection target.

[0095] The target that receives the most votes is identified as a moving target and output as such.

[0096] The technical solution in this embodiment utilizes the relative speed between detected targets in the image and preset voting conditions to vote for the detected targets, outputting the target with the most votes as the moving target. This enables accurate identification of moving targets even under conditions of image jitter, target movement, and field of view shifts, reducing the false alarm rate in the detection of weak moving targets. Furthermore, the processing method in this solution does not involve image processing, thus requiring fewer computational resources and improving processing speed.

[0097] A third aspect of the present invention provides a device for determining moving targets in infrared images, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of the method for determining moving targets in infrared images as described in any one of the first aspects of the present invention.

[0098] It should be noted that the computer-readable storage medium described in this embodiment includes, but is not limited to, ROM, RAM, disk, or optical disk. The processor may be a mobile phone, computer, server, air conditioner, or network device, etc.

[0099] A fourth aspect of the present invention provides a computer-readable storage medium storing an information transmission implementation program, which, when executed by a processor, implements the steps of the method for determining moving targets in an infrared image as described in any one of the first aspects of the embodiment.

[0100] It should be noted that well-known methods, structures, and techniques are not shown in detail in this specification so as not to obscure the understanding of this specification.

Claims

1. A method of determining a moving object in an infrared image, characterized by, The method comprises: acquiring all detected targets in a current frame image, and defining each detected target as a reference target in turn, and calculating relative speeds of other detected targets relative to the reference target; if the relative speeds of the other detected targets relative to the reference target are all greater than a first speed threshold, voting for the reference target once; if only one of the other detected targets has a relative speed greater than a second speed threshold, voting for the other detected target once; counting votes of each detected target, and determining a detected target with the most votes as a moving target, and other detected targets as false alarms; the calculation of the relative speeds of the other detected targets relative to the reference target comprises: acquiring historical coordinate values of detected targets in each historical frame image; calculating the relative speeds of the other detected targets relative to the reference target based on the historical coordinate values.

2. The method of claim 1, wherein, the calculation of the relative speeds of the other detected targets relative to the reference target based on the historical coordinate values comprises: calculating a relative velocity of other detected targets with respect to the reference target based on equation 1 : Formula 1, wherein S1 represents a distance of the other detected target and the reference target when they are located in the same historical frame image for the first time, S2 represents a distance of the other detected target and the reference target in the current frame image, S1 and S2 are in pixel units, and t represents a frame number difference between the historical frame image and the current frame image.

3. The method of claim 1, wherein, the calculation of the relative speeds of the other detected targets relative to the reference target comprises: calculating distances of the other detected targets and the reference target, and calculating the relative speeds of the other detected targets relative to the reference target only when the distances are less than or equal to a preset distance.

4. The method of claim 1, wherein, the acquisition of the historical coordinate values of the detected targets in each historical frame image comprises: acquiring a second threshold number of historical frame images that are greater than a first threshold number of historical frame images from the current frame image; acquiring historical coordinate values of detected targets in each of the second threshold number of historical frame images.

5. The method of claim 1, wherein, The method further comprises: counting a continuous frame number of a detected target determined as a moving target, and determining the moving target as a real moving target only when the continuous frame number is greater than a frame number threshold.

6. The method of claim 1, wherein, The first speed threshold and the second speed threshold are equal, and are both 0.2 pixels / frame.

7. A device for determining moving targets in infrared images, characterized in that, The method comprises: a motion information extraction unit, configured to acquire all detected targets in a current frame image, and define each detected target as a reference target in turn, and calculate relative speeds of other detected targets relative to the reference target; the calculation of the relative speeds of the other detected targets relative to the reference target comprises: acquiring historical coordinate values of detected targets in each historical frame image; calculating the relative speeds of the other detected targets relative to the reference target based on the historical coordinate values. a vote counting unit, configured to vote according to the following rules: if the relative speeds of the other detected targets relative to the reference target are all greater than a first speed threshold, voting for the reference target once; if only one of the other detected targets has a relative speed greater than a second speed threshold, voting for the other detected target once; The voting statistics unit is further configured to count the votes of each detection target after the voting is completed, and determine the detection target with the most votes as the moving target, and other detection targets as false alarms.

8. A device for determining moving targets in infrared images, characterized in that, The computer program comprises the steps of the method for determining a moving target in an infrared image according to any one of claims 1 to 6. The computer readable storage medium stores an implementation program of information transmission, and the program, when executed by the processor, implements the steps of the method for determining a moving target in an infrared image according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores an implementation program of information transmission, and the program, when executed by the processor, implements the steps of the method for determining a moving target in an infrared image according to any one of claims 1 to 6.

Images