Non-uniform noise correction

Abstract

A non-uniform noise correction method, an electronic device, and a computer program product. After correcting an output of an infrared imaging device at a current temperature each time, a non-uniform noise fpn_M is learned on the basis of a correction result (101). On the basis of the learned non-uniform noise fpn_M, background data corresponding to a highest temperature value and background data corresponding to a lowest temperature value in a temperature interval where the temperature is located are adjusted (102).

Description

Non-uniform noise correction Technical Field

[0001] This application relates to infrared imaging technology, and in particular to non-uniform noise correction methods, electronic devices, and computer program products. Background Technology

[0002] Non-uniform noise correction refers to the correction of non-uniform noise generated by infrared imaging equipment in infrared image applications. This non-uniform noise includes noise introduced by the infrared detector itself within the infrared imaging equipment, as well as noise introduced by radiation from structural elements such as lenses within the infrared imaging equipment. The infrared imaging equipment may or may not include an infrared imaging shield (hereinafter referred to as a shield); this application does not specifically limit its scope.

[0003] The presence of non-uniform noise can affect the quality of infrared images acquired by infrared imaging equipment, which in turn can affect subsequent image processing based on infrared images, such as target recognition. Summary of the Invention

[0004] This application provides a non-uniform noise correction method, electronic equipment, and computer program product for achieving non-uniform noise correction.

[0005] This application provides a non-uniform noise correction method applied to an infrared imaging device. The method includes: after correcting the output X of the infrared imaging device at the current temperature T based on third background data corresponding to the current temperature T, learning non-uniform noise fpn_M based on the correction result; wherein, the third background data corresponding to the current temperature T is determined by fitting based on first background data and second background data; the first background data and the second background data are respectively the background data corresponding to the lowest temperature value and the background data corresponding to the highest temperature value in the temperature range of the current temperature T from the stored background-temperature data; the temperature range is determined based on the temperature interval between each adjacent temperature in the background-temperature data; and updating the first background data and the second background data in the background-temperature data based on the non-uniform noise fpn_M.

[0006] This application provides a non-uniform noise correction method applied to an infrared imaging device. The method includes: determining a third background data corresponding to the current temperature T based on a first background data, a second background data, and a learned non-uniform noise fpn_M fitting; the first background data and the second background data are respectively the background data corresponding to the lowest temperature value and the background data corresponding to the highest temperature value in the temperature range where the current temperature T is located from the stored background-temperature data; the temperature range is determined based on the temperature interval between each adjacent temperature in the background-temperature data; and correcting the output X of the infrared imaging device at the current temperature T based on the third background data.

[0007] This application provides a non-uniform noise correction method applied to an infrared imaging device. The method includes: determining a third background data corresponding to the current temperature T based on first background data and second background data; the first background data and the second background data are respectively the background data corresponding to the lowest temperature value and the background data corresponding to the highest temperature value in the temperature range where the current temperature T is located from the stored background-temperature data; the temperature range is determined based on the temperature interval between each adjacent temperature in the background-temperature data; and correcting the output X of the infrared imaging device at the current temperature T based on the third background data and the learned non-uniform noise fpn_M.

[0008] This application also provides an electronic device. The electronic device includes: a processor and a machine-readable storage medium; the machine-readable storage medium stores machine-executable instructions executable by the processor; the processor is used to execute the machine-executable instructions to implement the steps of the disclosed method.

[0009] This application also provides a computer program product, which stores a computer program that, when executed by a processor, implements the steps of the method disclosed above.

[0010] As can be seen from the above technical solutions, in the embodiments of this application, after each correction of the output of the infrared imaging device at the current temperature, non-uniform noise fpn_M is learned based on the correction result, and the background data corresponding to the highest temperature value and the lowest temperature value in the temperature range of the current temperature are adjusted according to the learned non-uniform noise fpn_M. This method of updating the background data in real time with the dynamically learned non-uniform noise data can realize the real-time update of the background data in the stored background-temperature data based on the non-uniform noise data. Since the stored background-temperature data is used to correct the output of the infrared imaging device at a certain temperature, this is equivalent to realizing the non-uniform noise correction of the output of the infrared imaging device at that temperature.

[0011] Furthermore, in this embodiment, the method of updating the background data with the non-uniform noise data learned in real time ensures that even if the infrared imaging device switches from a moving state to a stationary state or is turned on statically, the output of the infrared imaging device can still be corrected using the background data updated based on the non-uniform noise data. This ensures the image quality throughout the moving and stationary processes, avoids the problem of poor image quality when the device is first turned on, and avoids the defect of incompatibility of the pre-stored background data after the infrared imaging device has undergone high and low temperature aging.

[0012] Furthermore, when fitting the third background data corresponding to the current temperature T each time, the non-uniform noise fpn_M is taken into account. The non-uniform noise fpn_M is included in determining the third background data corresponding to the current temperature T, ensuring that the final determined third background data is corrected by the non-uniform noise fpn_M. Then, the third background data corrected by the non-uniform noise fpn_M is used to correct the output X of the infrared imaging device at the current temperature T. This is equivalent to performing non-uniform noise correction on the output X of the infrared imaging device at the current temperature T.

[0013] Furthermore, by directly utilizing the learned non-uniform noise fpn_M to correct the output X of the infrared imaging device at the current temperature T, the correction of non-uniform noise is achieved. Attached Figure Description

[0014] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0015] Figure 1 is a flowchart of the method provided in an embodiment of this application.

[0016] Figure 2 is a flowchart of another method provided in an embodiment of this application.

[0017] Figure 3 is a flowchart of another method provided in an embodiment of this application.

[0018] Figures 4A and 4B are comparison diagrams of the effects provided by the embodiments of this application.

[0019] Figures 5A and 5B are another effect comparison diagrams provided by the embodiments of this application.

[0020] Figure 6 is a structural diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0021] To enable those skilled in the art to better understand the technical solutions provided in the embodiments of this application, and to make the above-mentioned objectives, features and advantages of the embodiments of this application more apparent and understandable, the technical solutions in the embodiments of this application will be further described in detail below with reference to the accompanying drawings.

[0022] The following describes how to obtain background temperature data.

[0023] As an example, the infrared imaging device can be placed in a high-low temperature chamber beforehand, and its lens can be covered with black foam. The infrared imaging device is then triggered to acquire infrared images at certain temperature intervals, such as every 5 degrees Celsius, to obtain infrared images (also known as baseline data) at each temperature. For example, the baseline data Base1 is obtained at temperature T1, the baseline data Base2 is obtained at temperature T2, and so on, eventually resulting in a series of baseline-temperature data: [Base1, Base2, ..., BaseN] and [T1, T2, ..., TN].

[0024] It should be noted that, in order to reduce the influence of random noise, the background data at any of the above temperatures can be determined by multi-frame calculation, such as by averaging. For example, taking the background data Base1 at temperature T1 as an example, multiple infrared images acquired by the infrared imaging device at temperature T1 can be averaged (for example, the gray values ​​of pixels at the same position in multiple infrared images can be averaged), and the result of the averaging can be used as the background data Base1 at temperature T1.

[0025] In the initial stage, for example, during the first calibration stage before the infrared imaging device leaves the factory, the aforementioned series of background-temperature data will be stored in the core of the infrared imaging device, such as the infrared imaging device itself.

[0026] Based on the above description, the method provided in the embodiments of this application will be described below.

[0027] Referring to Figure 1, which is a flowchart of a first method provided in an embodiment of this application, this method is applied to an infrared imaging device. As shown in Figure 1, the method may include steps 101-102.

[0028] Step 101: Correct the output X of the infrared imaging device at the current temperature T based on the third background data corresponding to the current temperature T, and learn the non-uniform noise fpn_M based on the correction result.

[0029] In this embodiment, before executing step 101, the first background data corresponding to the lowest temperature value and the second background data corresponding to the highest temperature value in the temperature range of the current temperature T can be determined from the stored background-temperature data. Based on the temperature intervals between adjacent temperatures in the background-temperature data described above, such as 5 degrees Celsius, each temperature range is implicitly included, such as 0-5 degrees Celsius, 5-10 degrees Celsius, and so on. Based on this, this embodiment can easily determine the temperature range of the current temperature T from the temperature ranges implicit in the stored background-temperature data. Then, the first background data corresponding to the lowest temperature value and the second background data corresponding to the highest temperature value can be found from the stored background-temperature data.

[0030] As an example, the first baseline data and the second baseline data are the baseline data obtained in the initial stage described above, or they can be the baseline data that has been updated in a manner similar to step 102 below. This example is not specifically limited.

[0031] After obtaining the first background data corresponding to the lowest temperature value and the second background data corresponding to the highest temperature value, a fitting method can be used to determine the third background data corresponding to the current temperature T. There are many fitting methods, such as linear fitting and nonlinear fitting.

[0032] Taking linear fitting as an example, assuming that the first background data corresponding to the lowest temperature value (denoted as T_L) is B_L, and the second background data corresponding to the highest temperature value (denoted as T_H) is B_H, then the third background data corresponding to the current temperature T (denoted as B_ins) can be determined as follows: B_ins=B_H*(T–T_L) / (T_H–T_L)+B_L*(T_H–T) / (T_H–T_L).

[0033] The output X of the infrared imaging device at the current temperature T is corrected based on the third background data corresponding to the current temperature T. There are many ways to implement this, such as by using the following formula: Y=K*(X–B_ins)+Offset.

[0034] Where Y represents the correction result, and K and Offset are the set gain coefficient and set offset, respectively. B_ins are as described above. X represents the output of the infrared imaging device at the current temperature T. Optionally, X–B_ins can represent the calculation of the difference in grayscale values ​​of pixels at the same position corresponding to X and B_ins, respectively.

[0035] In this embodiment, the non-uniform noise fpn_M can be extracted using a scene-based online learning scheme, or it can be extracted using methods such as temporal filtering, constant statistics, registration, neural networks, etc. This embodiment is not specifically limited. However, the overall idea is the same: motion judgment, noise learning and extraction, and accumulation to obtain the non-uniform noise fpn_M. It should be noted that the non-uniform noise may be different in different learning iterations, and this embodiment is not specifically limited.

[0036] Step 102: Update the first and second background data in the stored background-temperature data based on the non-uniform noise fpn_M.

[0037] As an example, updating the first background data in the background-temperature data based on non-uniform noise fpn_M includes: substituting the first background data, the current temperature T, the highest temperature value, and the lowest temperature value into a first specified algorithm to obtain the updated first background data. For example, the first background data is updated according to the following formula: B_L'=B_L+fpn_M*(T–T_L) / (T_H–T_L).

[0038] Where B_L' represents the updated first background data, T_L represents the aforementioned minimum temperature value, T_H represents the aforementioned maximum temperature value, and B_L represents the aforementioned first background data.

[0039] As an example, updating the second background data in the background-temperature data based on non-uniform noise fpn_M includes: substituting the second background data, the current temperature T, the aforementioned highest temperature value, and the aforementioned lowest temperature value into a second specified algorithm to obtain the updated second background data. For example, the second background data is updated according to the following formula: B_H'=B_H+fpn_M*(T_H-T) / (T_H–T_L).

[0040] Where B_L' represents the updated second background data, T_L represents the aforementioned minimum temperature value, T_H represents the aforementioned maximum temperature value, and B_H represents the aforementioned second background data.

[0041] It can be observed that in this embodiment, after each correction of the output of the infrared imaging device at the current temperature, non-uniform noise fpn_M is learned based on the correction result. The second background data corresponding to the highest temperature value and the first background data corresponding to the lowest temperature value in the temperature range are adjusted according to the learned non-uniform noise fpn_M. This method of updating the background data with dynamically learned non-uniform noise data in real time can directly use the background data updated based on non-uniform noise data to correct the output of the infrared imaging device in static scenarios, especially when the infrared imaging device is statically powered on. This avoids the problem of poor image quality when the device is first powered on and also avoids the defect of incompatibility of the pre-stored background data after the infrared imaging device has undergone high and low temperature aging.

[0042] Furthermore, this method of updating the non-uniform noise data learned dynamically into the background data in real time means that the stored background data is updated in real time based on the non-uniform noise. It also means that the output of the infrared imaging device at a certain temperature is corrected based on the updated background data. This allows the method of learning non-uniform noise based on the correction results to accelerate the convergence speed of non-uniform noise.

[0043] This completes the description of the process shown in Figure 1.

[0044] Another embodiment of this application is described below.

[0045] Referring to Figure 2, which is another method flowchart provided in an embodiment of this application, this method is applied to an infrared imaging device. As shown in Figure 2, the method includes steps 201-202.

[0046] Step 201: Based on the first background data, the second background data, and the learned non-uniform noise fpn_M, determine the third background data corresponding to the current temperature T. The first background data and the second background data are the background data corresponding to the lowest temperature value and the background data corresponding to the highest temperature value in the temperature range where the current temperature T is located, respectively, from the stored background-temperature data. The temperature range is determined based on the temperature interval between each adjacent temperature in the stored background-temperature data.

[0047] For example, in this embodiment, the first background data corresponding to the lowest temperature value and the second background data corresponding to the highest temperature value within the temperature range of the current temperature T are first determined from the stored background-temperature data. As described above, the background-temperature data, based on the temperature intervals between adjacent temperatures (e.g., 5 degrees Celsius), implicitly contains temperature ranges such as 0-5 degrees Celsius, 5-10 degrees Celsius, and so on. Based on this, this embodiment can easily determine the temperature range of the current temperature T. Then, the first background data corresponding to the lowest temperature value and the second background data corresponding to the highest temperature value within the temperature range of the current temperature T can be found from the stored background-temperature data.

[0048] Then, the third background data corresponding to the current temperature T can be determined by fitting the first background data, the second background data, and the learned non-uniform noise fpn_M. For example, based on the aforementioned minimum temperature value and the aforementioned maximum temperature value, the first background data and the second background data are fitted to obtain the fitting result corresponding to the current temperature T; there are many ways to do this fitting, such as linear fitting and nonlinear fitting. Taking linear fitting as an example, assuming that the first background data corresponding to the aforementioned minimum temperature value (denoted as T_L) is B_L, and the second background data corresponding to the maximum temperature value (denoted as T_H) is B_H, then the fitting result can be determined as follows: B_ins'=B_H*(T–T_L) / (T_H–T_L)+B_L*(T_H–T) / (T_H–T_L).

[0049] Then, based on the difference between the fitting result and the non-uniform noise fpn_M, the third background data is determined. For example, the third background data (denoted as B_ins) corresponding to the current temperature T can be: B_ins = B_ins' - fpn_M.

[0050] It should be noted that the non-uniform noise fpn_M here can be the non-uniform noise learned from the previous correction result. The learning method is as described above, and this embodiment is not specifically limited.

[0051] Step 202: Correct the output X of the infrared imaging device at the current temperature T based on the third background data.

[0052] As for how to correct the output X of the infrared imaging device at the current temperature T based on the third background data corresponding to the current temperature T, there are many ways to implement it, such as correcting it according to the following formula: Y=K*(X–B_ins)+Offset.

[0053] Where Y represents the correction result, and K and Offset are the set gain coefficient and set offset, respectively. B_ins are as described above. X represents the output of the infrared imaging device at the current temperature T. Optionally, X–B_ins can represent the calculation of the difference in grayscale values ​​of pixels at the same position corresponding to X and B_ins, respectively.

[0054] This completes the process shown in Figure 2.

[0055] As can be seen from the flowchart shown in Figure 2, in this embodiment, when fitting the third background data corresponding to the current temperature T each time, the non-uniform noise fpn_M is taken into account. The non-uniform noise fpn_M is included in determining the third background data corresponding to the current temperature T, ensuring that the finally determined third background data is corrected by the non-uniform noise fpn_M. Then, the third background data corrected by the non-uniform noise fpn_M is used to correct the output X of the infrared imaging device at the current temperature T, which is equivalent to realizing the non-uniform noise correction of the output X of the infrared imaging device at the current temperature T.

[0056] Another embodiment of this application is described below.

[0057] Referring to Figure 3, which is another method flowchart provided in an embodiment of this application, this method is applied to an infrared imaging device. As shown in Figure 3, the process may include steps 301-302.

[0058] Step 301: Based on the first background data and the second background data, determine the third background data corresponding to the current temperature T by fitting the data. The first background data and the second background data are the background data corresponding to the lowest temperature value and the background data corresponding to the highest temperature value in the temperature range where the current temperature T is located, respectively, from the stored background-temperature data. The temperature range is determined based on the temperature interval between each adjacent temperature in the stored background-temperature data.

[0059] For example, in this embodiment, the first background data corresponding to the lowest temperature value and the second background data corresponding to the highest temperature value within the temperature range of the current temperature T are first determined from the stored background-temperature data. Based on the temperature intervals between adjacent temperatures in the background-temperature data described above, such as 5 degrees Celsius, the temperature ranges are implicitly defined, such as 0-5 degrees Celsius, 5-10 degrees Celsius, and so on. Based on this, the temperature range of the current temperature T can be easily determined in this embodiment. Then, the first background data corresponding to the lowest temperature value and the second background data corresponding to the highest temperature value within the temperature range of the current temperature T can be found from the stored background-temperature data.

[0060] Then, the third background data corresponding to the current temperature T can be determined by fitting the first and second background data. For example, based on the highest and lowest temperature values ​​mentioned above, the first and second background data can be fitted to obtain the third background data corresponding to the current temperature T. There are many ways to perform this fitting, such as linear fitting and nonlinear fitting. Taking linear fitting as an example, assuming that the first background data corresponding to the lowest temperature value (denoted as T_L) is B_L, and the second background data corresponding to the highest temperature value (denoted as T_H) is B_H, then the third background data corresponding to the current temperature T (denoted as B_ins) can be determined as follows: B_ins = B_H*(T–T_L) / (T_H–T_L) + B_L*(T_H–T) / (T_H–T_L).

[0061] Step 302: Based on the third background data and the learned non-uniform noise fpn_M, the output X of the infrared imaging device at the current temperature T is corrected.

[0062] In this embodiment, the correction of the output X of the infrared imaging device at the current temperature T based on the third background data and the learned non-uniform noise fpn_M may include: substituting the above-mentioned output X, the third background data, the set gain coefficient K and the set offset into the third specified algorithm to obtain the preliminary correction result of the output X (denoted as Y'), for example: Y'=K*(X–B_ins)+Offset.

[0063] Then, the preliminary correction result is corrected based on the learned non-uniform noise fpn_M to obtain the target correction result. For example, the target correction result (denoted as Y) can be expressed by the following formula: Y = K*(X – B_ins) – fpn_M + Offset.

[0064] As can be seen, in this embodiment, the output X of the infrared imaging device at the current temperature T is directly corrected using the learned non-uniform noise fpn_M, thus achieving non-uniform noise correction. It should be noted that the non-uniform noise fpn_M here can be the non-uniform noise learned from the previous correction result. The learning method is described above, and this embodiment is not specifically limited to it.

[0065] This completes the process shown in Figure 3.

[0066] Figures 4A and 4B show a comparison of the effects before and after correction using any of the methods described above. Figures 5A and 5B show a comparison of the effects before and after correction using any of the methods described above. It can be seen that correction using any of the methods described above can greatly improve the quality of infrared images acquired by infrared imaging devices, thereby improving the accuracy of subsequent image processing based on infrared images, such as target recognition.

[0067] The proposed solution is particularly suitable for barrier-free thermal imaging devices. Barriers are the structure used in traditional thermal imaging modules to acquire noise correction data. During the imaging process, the module needs to be triggered periodically, which can cause scene freezing. Barrier-free devices eliminate the barrier structure, allowing for a smaller module size and lower power consumption, and eliminating the image freezing problem.

[0068] The methods provided in the embodiments of this application have been described above. The apparatus provided in the embodiments of this application will be described below.

[0069] Referring to Figure 6, which is a structural diagram of an electronic device provided in an embodiment of this application, the hardware structure may include a processor and a machine-readable storage medium. The machine-readable storage medium stores machine-executable instructions that can be executed by the processor. The processor is used to execute the machine-executable instructions to implement the method disclosed in the above example of this application.

[0070] Based on the same application concept as the above method, this application embodiment also provides a machine-readable storage medium storing a plurality of computer instructions, which, when executed by a processor, can implement the method disclosed in the above examples of this application.

[0071] Based on the same application concept as the above method, this application embodiment also provides a computer program product storing a computer program, which, when executed by a processor, implements the method disclosed in the above examples of this application.

[0072] For example, the aforementioned machine-readable storage medium can be any electronic, magnetic, optical, or other physical storage device that can contain or store information such as executable instructions, data, etc. For instance, machine-readable storage media can be: RAM (Random Access Memory), volatile memory, non-volatile memory, flash memory, storage drives (such as hard disk drives), solid-state drives, any type of storage disk (such as optical discs, DVDs, etc.), or similar storage media, or combinations thereof.

[0073] The above descriptions are merely some embodiments of this application and are not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A non-uniform noise correction method, applied to an infrared imaging device, the method comprising: The third background data corresponding to the current temperature T is determined by fitting the first background data and the second background data in the stored background-temperature data. The first background data is the background data corresponding to the lowest temperature value in the temperature range where the current temperature T is located, and the second background data is the background data corresponding to the highest temperature value in the temperature range where the current temperature T is located. The temperature range is determined based on the temperature interval between each adjacent temperature in the background-temperature data. The output X of the infrared imaging device at the current temperature T is corrected based on the third background data corresponding to the current temperature T. Based on the correction results, learn the non-uniform noise fpn_M; The first background data and the second background data in the background-temperature data are updated based on the non-uniform noise fpn_M.

2. The method according to claim 1, wherein, The step of updating the first background data in the background-temperature data based on the non-uniform noise fpn_M includes: substituting the first background data, the current temperature T, the highest temperature value, and the lowest temperature value into a first specified algorithm to obtain the updated first background data; The step of updating the second background data in the background-temperature data based on the non-uniform noise fpn_M includes: substituting the second background data, the current temperature T, the highest temperature value, and the lowest temperature value into a second specified algorithm to obtain the updated second background data.

3. The method of claim 2, wherein, The first specified algorithm is: B_L' = B_L + fpn_M * (T - T_L) / (T_H - T_L); Wherein, B_L' represents the updated first background data. T_L represents the minimum temperature value. T_H represents the highest temperature value. B_L represents the first background data.

4. The method of claim 2, wherein, The second specified algorithm is: B_H' = B_H + fpn_M * (T_H - T) / (T_H - T_L); Wherein, B_H' represents the updated second background data. T_L represents the minimum temperature value. T_H represents the highest temperature value. B_H represents the second background data.

5. The method according to claim 1, further comprising: Based on the first background data, the second background data, and the learned non-uniform noise fpn_M, the updated third background data corresponding to the current temperature T is determined; The output X of the infrared imaging device at the current temperature T is corrected based on the updated third background data.

6. The method of claim 1, wherein, The step of determining the updated third background data corresponding to the current temperature T based on the first background data, the second background data, and the learned non-uniform noise fpn_M includes: Based on the lowest and highest temperature values ​​within the temperature range of the current temperature T, the first background data and the second background data are fitted to obtain a fitting result. Based on the difference between the fitting result and the non-uniform noise fpn_M, the updated third background data is determined.

7. The method according to claim 1, further comprising: Based on the third background data and the learned non-uniform noise fpn_M, the output X of the infrared imaging device at the current temperature T is corrected.

8. The method of claim 7, wherein, The step of correcting the output X of the infrared imaging device at the current temperature T based on the third background data and the learned non-uniform noise fpn_M includes: Substitute the output X, the third background data, the set gain coefficient K, and the set offset into the third specified algorithm to obtain the preliminary correction result of the output X; The preliminary correction result is corrected based on the learned non-uniform noise fpn_M to obtain the target correction result.

9. An electronic device, comprising: The electronic device includes: a processor and a machine-readable storage medium; The machine-readable storage medium stores machine-executable instructions that can be executed by the processor; The processor is configured to execute machine-executable instructions to implement the method steps of any one of claims 1-8.

10. A computer program product, wherein, The computer program product contains a computer program that, when executed by a processor, implements the method of any one of claims 1 to 8.

Images