Low-frequency noise blind pixel detection method and system based on segmented noise difference

Abstract

This invention discloses a low-frequency noise blind pixel detection method and system based on segmented noise difference. The method involves observing all pixels of the detector against a uniform and stable blackbody to obtain observation data for each pixel (f represents the frame number); calculating the noise of each pixel's observation data; segmenting the observation data and then combining them; calculating the noise of each combined new data segment; and calculating the first low-frequency stability index LS1 for each pixel based on the noise of each pixel's observation data and the noise of each combined new data segment. i Second Low Frequency Stability Index LS2 i The first low-frequency stability index LS1 i Below the first low-frequency stability index threshold and the second low-frequency stability index LS2 i Pixels below the second low-frequency stability index threshold are identified as low-frequency noise blind pixels. This effectively avoids missing detection of low-frequency noise blind pixels.

Description

Technical Field

[0001] This invention relates to the field of infrared detector signal processing technology, and in particular to a method and system for detecting low-frequency noise blind elements based on segmented noise differences. Background Technology

[0002] With the continuous development of aerospace technology, mobile vehicles such as aircraft pods, drones, and other aircraft are increasingly widely used in modern military reconnaissance, civilian surveying and mapping, and environmental monitoring. Infrared cameras mounted on these vehicles, capable of operating normally at night or in adverse weather conditions, have become an important means of acquiring target information. However, during the imaging process, infrared cameras are prone to blind pixels due to various factors such as unstable internal component performance and external environmental interference. Blind pixels, which are points with abnormal or missing pixel values ​​in an image, severely affect the clarity and accuracy of infrared images, reduce the reliability of target identification, and may even lead to the loss of critical information.

[0003] The problem of dead pixels in infrared imaging systems manifests as a complex non-uniformity. The causes of dead pixels primarily encompass issues with the device itself, such as pixel damage and inconsistent responsivity, as well as signal communication problems, such as focal plane readout circuitry and communication obstacles. The causes of dead pixels are diverse and complex, and regardless of their form, they will affect the imaging quality of infrared images to some extent. Therefore, the handling of dead pixels is particularly important.

[0004] In GB / T 17444-2013, the criterion for a blind pixel is that its responsivity is less than half of the average responsivity, and the criterion for a noisy pixel is that its noise voltage is greater than twice the average noise voltage. In engineering, signal values ​​and sensitivity are also commonly used for judgment. After removing blind pixels, subsequent processing and applications are carried out.

[0005] Low-frequency noise refers to noise with energy concentrated in the low-frequency region (less than 10Hz), often manifesting as signal value drift, fluctuation, and jumps. The traditional blind pixel detection methods mentioned above have poor accuracy for this type of low-frequency noise, failing to promptly identify low-frequency noise blind pixels as blind pixels, further affecting the detection of subsequent blind pixels and severely impacting data usability. Therefore, there is an urgent need for a low-frequency noise detection method with high accuracy and low complexity. Summary of the Invention

[0006] To address the aforementioned problems, the present invention aims to provide a method and system for detecting low-frequency noise blind elements based on segmented noise differences, which can effectively avoid missed detection of low-frequency noise blind elements.

[0007] The technical solution provided by this invention is: a low-frequency noise blind cell detection method based on segmented noise differences, comprising the following steps:

[0008] The detector's pixels are observed against a uniform and stable blackbody to obtain observation data for each pixel. f represents the frame number;

[0009] Calculate the observation data for each pixel noise

[0010] The observation data After segmenting, the data is combined, and the noise of each new combined data segment is calculated.

[0011] Based on observation data of each pixel noise and the noise in each new data segment after combination Calculate the first low-frequency stability index LS1 for each pixel. i Second Low Frequency Stability Index LS2 i ;

[0012] The first low-frequency stability index LS1 i Below the first low-frequency stability index threshold and the second low-frequency stability index LS2 i Pixels below the second low-frequency stability index threshold are classified as low-frequency noise blind pixels.

[0013] Preferably, the observation data Segmentation followed by recombination further includes:

[0014] Observation data The frame is divided into N segments, which are connected end-to-end, according to the frame number f. Where part is the segment number, ranging from 1 to N, and f part The frame number for each data segment;

[0015] Combine any two consecutive data segments end-to-end to form new data. In the formula The new data after combination, where P is the segment number of the new data after combination, taking a value from 1 to N-1, f P This is the frame number for each new data segment.

[0016] Preferably, the noise of each new data segment after calculation and combination is calculated. Further includes:

[0017] Calculate each new data segment noise

[0018]

[0019] In the formula For data The average value, f P for Frame number, F P for Total number of frames.

[0020] Preferably, the method further includes discarding the remainder or making the number of frames in each data segment inconsistent when F cannot divide N.

[0021] Preferably, based on observation data of each pixel noise The noise in each new data segment after combination Calculate the first low-frequency stability index LS1 for each pixel. i Second Low Frequency Stability Index LS2 i Further includes:

[0022] Calculate the first low-frequency stability index LS1 for each pixel. i :

[0023]

[0024] Calculate the second low-frequency stability index LS2 for each pixel. i :

[0025]

[0026] In the formula, represent The minimum value in, represent The maximum value in.

[0027] Preferably, the first low-frequency stability index threshold is a preset value or the first low-frequency stability index LS1 of all pixels. i The threshold value is 0.5 times the second low-frequency stability index threshold, which is either a preset value or the second low-frequency stability index LS2 for all pixels. i 0.5 times.

[0028] Preferably, it is applied to the ground testing phase of the optical payload detector of the remote sensing satellite, or to its on-orbit operation phase; when the instrument is in orbit, during the instrument's idle phase, an observation of the spaceborne blackbody for no less than 30 seconds is arranged daily or weekly to update the list of low-frequency noise blind elements.

[0029] Based on the same concept, the present invention also provides a low-frequency noise blind cell detection system based on segmented noise differences, comprising:

[0030] The data acquisition module is used to observe all pixels of the detector against a uniform and stable blackbody to obtain observation data for each pixel. f represents the frame number;

[0031] The preliminary calculation module is used to calculate the observation data for each pixel. noise

[0032] The recombination module is used to reconstruct the observation data. After segmenting, the data is combined, and the noise of each new combined data segment is calculated.

[0033] The index calculation module is used to calculate the index based on the observation data of each pixel. noise and the noise in each new data segment after combination Calculate the first low-frequency stability index LS1 for each pixel. i Second Low Frequency Stability Index LS2 i ;

[0034] The blind cell determination module is used to determine the first low-frequency stability index LS1. i Below the first low-frequency stability index threshold and the second low-frequency stability index LS2 i Pixels below the second low-frequency stability index threshold are classified as low-frequency noise blind pixels.

[0035] Based on the same concept, the present invention also provides an electronic device, comprising:

[0036] The memory is used to store the processing program;

[0037] A processor, which, when executing the processing program, implements the low-frequency noise blind cell detection method based on segmented noise differences as described above.

[0038] Based on the same concept, the present invention also provides a readable storage medium storing a processing program, which, when executed by a processor, implements the low-frequency noise blind cell detection method based on segmented noise difference described above.

[0039] Because the present invention adopts the above technical solution, it has the following advantages and positive effects compared with the prior art:

[0040] Before using traditional blind pixel identification methods, the technical solution of this invention also includes observation data. After segmenting, the data is combined, and the noise of each new combined data segment is calculated. Based on the observation data of each pixel noise and the noise in each new data segment after combination Calculate the first low-frequency stability index LS1 for each pixel. i Second Low Frequency Stability Index LS2i This allows for the identification of low-frequency noise blind cells based on threshold values, effectively preventing missed detections of low-frequency noise blind cells. Attached Figure Description

[0041] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings, wherein:

[0042] Figure 1 This is a flowchart of the low-frequency noise blind cell detection method based on segmented noise differences of the present invention;

[0043] Figure 2 Observation data in the embodiments of the present invention A diagram illustrating the process of segmenting and then combining the segments;

[0044] Figure 3 This refers to the drift characteristics of low-frequency noise.

[0045] Figure 4 This refers to the fluctuation characteristics of low-frequency noise.

[0046] Figure 5 This refers to the jumping characteristics of low-frequency noise. Detailed Implementation

[0047] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The advantages and features of the present invention will become clearer from the following description and claims. It should be noted that the drawings are all in a very simplified form and use non-precise ratios, and are only used to facilitate and clarify the illustration of the embodiments of the present invention.

[0048] It should be noted that all directional indications in the embodiments of the present invention (such as up, down, left, right, front, back, etc.) are only used to explain the relative position relationship, movement status, etc. between the various components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0049] First Embodiment

[0050] The detection scheme proposed in this invention can be applied to both the ground testing phase and the on-orbit operation phase of remote sensing satellite optical payload detectors. The low-frequency noise detection method will be described in detail below, taking the ground testing phase as an example.

[0051] like Figure 1 As shown, a low-frequency noise blind cell detection method based on segmented noise differences includes the following steps:

[0052] The detector's pixels are observed against a uniform and stable blackbody to obtain observation data for each pixel. f represents the frame number;

[0053] Calculate the observation data for each pixel noise

[0054] The observation data After segmenting, the data is combined, and the noise of each new combined data segment is calculated.

[0055] Based on observation data of each pixel noise The noise in each new data segment after combination Calculate the first low-frequency stability index LS1 for each pixel. i Second Low Frequency Stability Index LS2 i ;

[0056] The first low-frequency stability index LS1 i Below the first low-frequency stability index threshold and the second low-frequency stability index LS2 i Pixels below the second low-frequency stability index threshold are classified as low-frequency noise blind pixels.

[0057] The existing blind pixel detection schemes used in engineering for infrared optical payload detectors of remote sensing satellites are as follows:

[0058] The detector observes a uniform and stable blackbody (temperature T1) for a period of time (more than 2 seconds, more than 1000 frames), and calculates the average value S of the signal from each detector pixel. i (T1), standard deviation σ i (T1), standard deviation σ i (T1) is noise. i (T1)

[0059] The detector observes a uniform and stable blackbody (temperature T2, approximately 10K different from T1) for a period of time (more than 2 seconds, more than 1000 frames), and calculates the average value S of the signal from each detector pixel. i (T2), standard deviation σ i (T2), standard deviation σ i (T2) is noise. i (T2)

[0060] Based on the above data, the noise equalization temperature difference NEdT under T1 can be approximated. i and response rate R i

[0061]

[0062] R i =S i (T1)-S i (T2)

[0063] Based on the above calculation results, the signal values ​​S are sequentially... i (T1) Pixels that do not meet the requirement of being within 0.5 to 1.5 times the average signal value of all pixels, and whose responsivity R i Pixels that do not meet the requirement of having a response rate between 0.5 and 1.5 times the average response rate of all pixels, and noise. i (T1) Pixels with noise equivalent temperature difference NEDT greater than twice the average noise of all pixels. i Pixels with a temperature (T1) greater than the index requirement are identified as blind pixels at temperature T1.

[0064] This method is simple and easy to implement, but it may miss low-frequency noise blind elements. Low-frequency noise refers to noise whose energy is mainly concentrated in the low-frequency region (below 10Hz), typically manifested as signal value drift, fluctuation, and jumps. Traditional blind element detection methods have poor accuracy for this type of low-frequency noise, failing to identify low-frequency noise blind elements in a timely manner, further affecting the detection of subsequent blind elements and severely impacting the usability of remote sensing data.

[0065] Existing technologies directly calculate the noise of each pixel, and low-frequency noise may not appear during short-term (2s) observations. However, even with increased observation time, traditional detection methods still cannot accurately identify detector pixels with low-frequency noise. This is because when low-frequency noise exists in the detector, its overall signal value S... i Response rate R i noise i Noise equalization temperature difference NEDT i It is still possible that the pixel may fall within the normal pixel range, making it impossible to accurately identify it as a blind pixel, and further affecting the blind pixel detection results of other pixels. The technical solution of this embodiment, before using traditional blind pixel identification methods, also considers the observed data... After segmenting, the data is combined, and the noise of each new combined data segment is calculated. Based on the observation data of each pixel noise The noise in each new data segment after combination Calculate the first low-frequency stability index LS1 for each pixel. i Second Low Frequency Stability Index LS2 i This allows for the identification of low-frequency noise blind cells based on threshold values, effectively preventing missed detections. Furthermore, the algorithm boasts low complexity and fast detection speed.

[0066] Preferably, the observation data Segmentation followed by recombination further includes:

[0067] Observation data The frame is divided into N segments, which are connected end-to-end, according to the frame number f. Where part is the segment number, ranging from 1 to N, and f part The frame number for each data segment;

[0068] Combine any two consecutive data segments end-to-end to form new data. In the formula The new data after combination, where P is the segment number of the new data after combination, taking a value from 1 to N-1, f P This is the frame number for each new data segment.

[0069] For example, see Figure 2 The observation data is shown. A diagram illustrating the segmentation and subsequent recombination of observation data. The frame is divided into N segments (where N is a positive integer, generally greater than 10) based on its sequence number f, and these segments are denoted as follows: Where part is the segment number, ranging from 1 to N, and f part The frame number for each data segment, i.e. Divided into There are N segments in total. (When F is not divisible by N, the remainder can be discarded, or the number of frames in each segment can be made inconsistent). Each pair of consecutive segments is concatenated end-to-end to form new data, specifically as follows:

[0070]

[0071] In the formula The new data after combination, where P is the segment number of the new data after combination, taking a value from 1 to N-1, f P [a,b] represents the frame number of each new data segment, where [a,b] represents combining data a and b end-to-end to form a new data segment.

[0072] Preferably, the noise of each new data segment after calculation and combination is calculated. Further includes:

[0073] Calculate each new data segment noise

[0074]

[0075] In the formula For data The average value, f P for Frame number, F P for Total number of frames.

[0076] For example, see Figures 3-5 The drift, fluctuation, and jump characteristics of low-frequency noise are shown respectively. Figure 3 and Figure 4 In the data, we select data 1 and calculate its standard deviation. The standard deviation of data 1 is much smaller than the standard deviation of all data. Therefore, the value of standard deviation (data 1) / standard deviation (all data) is very small. The calculated first stability index is also very small. Therefore, we can determine that there is drift or low-frequency noise in the data. Figure 5 In the data, the regions of data 1 and data 2 are selected, and their standard deviations are calculated. Obviously, the standard deviation of data 1 is much smaller than that of data 2. Therefore, the value of standard deviation (data 1) / standard deviation (data 2) is very small, and the calculated second stability index is very small, thus indicating that there is low-frequency noise in the data.

[0077] Preferably, the method further includes discarding the remainder or making the number of frames in each data segment inconsistent when F cannot divide N.

[0078] In some cases, even if F cannot divide N, it is still necessary to maintain data consistency and integrity. In this case, by adjusting the number of frames in each data segment (i.e., not distributing data equally) or discarding the remainder, it is possible to achieve a uniform distribution or a specific format arrangement of data while keeping the overall data volume unchanged.

[0079] Preferably, based on observation data of each pixel noise The noise in each new data segment after combination Calculate the first low-frequency stability index LS1 for each pixel. i Second Low Frequency Stability Index LS2 i Further includes:

[0080] Calculate the first low-frequency stability index LS1 for each pixel. i :

[0081]

[0082] Calculate the second low-frequency stability index LS2 for each pixel. i :

[0083]

[0084] In the formula, represent The minimum value in, represent The maximum value in.

[0085] The first low-frequency stability index LS1 is calculated. i Second Low Frequency Stability Index LS2i The higher the values ​​of both, the lower the low-frequency noise and the better the low-frequency stability of the pixel.

[0086] Preferably, the first low-frequency stability index threshold is the first low-frequency stability index LS1 of all pixels. i The threshold for the second low-frequency stability index is 0.5 times that of the second low-frequency stability index LS2 for all pixels. i 0.5 times.

[0087] The first low-frequency stability index LS1 i Below the first low-frequency stability index threshold and the second low-frequency stability index LS2 i Pixels with values ​​below the second low-frequency stability index threshold are classified as low-frequency noise blind pixels. The first and second low-frequency stability index thresholds can be set manually or defined as 0.5 times the average of the first and second low-frequency stability indices for all pixels, respectively.

[0088] Preferably, it is applied to the ground testing phase of the optical payload detector of the remote sensing satellite, or to its on-orbit operation phase; when the instrument is in orbit, during the instrument's idle phase, an observation of the spaceborne blackbody for no less than 30 seconds is arranged daily or weekly to update the list of low-frequency noise blind elements.

[0089] After performing the above method, low-frequency noise blind element detection is achieved. Then, blind element detection is performed according to the traditional method, which can effectively avoid the missed detection of low-frequency noise blind elements.

[0090] When the instrument is in orbit, the observation time for the spaceborne blackbody is about 2 seconds each time, which is insufficient for low-frequency noise detection. During the instrument's idle period, observations of the spaceborne blackbody for about 30 seconds can be arranged daily or weekly to update the list of low-frequency noise blind elements.

[0091] Based on the same concept, the present invention also provides a low-frequency noise blind cell detection system based on segmented noise differences, comprising:

[0092] The data acquisition module is used to observe all pixels of the detector against a uniform and stable blackbody to obtain observation data for each pixel. f represents the frame number;

[0093] The preliminary calculation module is used to calculate the observation data for each pixel. noise

[0094] The recombination module is used to reconstruct the observation data. After segmenting, the data is combined, and the noise of each new combined data segment is calculated.

[0095] The index calculation module is used to calculate the index based on the observation data of each pixel. noise The noise in each new data segment after combination Calculate the first low-frequency stability index LS1 for each pixel. i Second Low Frequency Stability Index LS2 i ;

[0096] The blind cell determination module is used to determine the first low-frequency stability index LS1. i Below the first low-frequency stability index threshold and the second low-frequency stability index LS2 i Pixels below the second low-frequency stability index threshold are classified as low-frequency noise blind pixels.

[0097] Based on the same concept, the present invention also provides an electronic device, comprising:

[0098] The memory is used to store the processing program;

[0099] A processor, which, when executing the processing program, implements the low-frequency noise blind cell detection method based on segmented noise differences as described above.

[0100] Based on the same concept, the present invention also provides a readable storage medium storing a processing program, which, when executed by a processor, implements the low-frequency noise blind cell detection method based on segmented noise difference described above.

[0101] The low-frequency noise blind element detection method based on segmented noise differences, if implemented as program instructions and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment, essentially, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in software. This computer software is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0102] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific identification content executed by the system and device described above can be referred to the corresponding process in the foregoing method embodiments.

[0103] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to the above embodiments. Even if various changes are made to the present invention, if these changes fall within the scope of the claims of the present invention and their equivalents, they shall still fall within the protection scope of the present invention.

Claims

1. A method for detecting low-frequency noise blind elements based on segmented noise differences, characterized in that, The following steps are involved: The detector's pixels are observed against a uniform and stable blackbody to obtain observation data for each pixel. f represents the frame number; Calculate the observation data for each pixel noise ; The observation data After segmenting, the data is combined, and the noise of each new combined data segment is calculated. ; wherein, the observation data The subsequent segmentation and recombination of observation data further includes: The frame is divided into N segments, which are connected end-to-end, according to the frame number f. Where part is the segment number, with a value from 1 to N. The frame number of each data segment; concatenate two consecutive data segments end-to-end to form a new data segment. In the formula The new data is generated after combination, where P is the segment number of the new data after combination, and its value ranges from 1 to N-1. Assign the frame number to each new data segment; calculate the noise of each combined new data segment. Further includes: calculating each new data segment noise : In the formula For data The average value, for Frame number, for Total number of frames; Based on observation data of each pixel noise and the noise in each new data segment after combination Calculate the first low-frequency stability index for each pixel. Second low-frequency stability index Further includes: calculating the first low-frequency stability index for each pixel. : Calculate the second low-frequency stability index for each pixel. : In the formula, represent , , The minimum value in, represent , , The maximum value in; The first low-frequency stability index Below the first low-frequency stability index threshold and the second low-frequency stability index Pixels below the second low-frequency stability index threshold are classified as low-frequency noise blind pixels.

2. The low-frequency noise blind cell detection method based on segmented noise difference according to claim 1, characterized in that, The method also includes discarding the remainder or making the number of frames in each data segment inconsistent when F cannot divide N.

3. The low-frequency noise blind cell detection method based on segmented noise difference according to claim 1, characterized in that, The first low-frequency stability index threshold is a preset value or the first low-frequency stability index of all pixels. The second low-frequency stability index threshold is 0.5 times the preset value or the second low-frequency stability index of all pixels. 0.5 times.

4. The low-frequency noise blind cell detection method based on segmented noise difference according to claim 1, characterized in that, It is applied to the ground testing phase of the optical payload detector of remote sensing satellite, or to its on-orbit operation phase; when the instrument is in orbit, during the instrument's idle phase, an observation of the spaceborne blackbody for no less than 30 seconds is scheduled daily or weekly to update the list of low-frequency noise blind elements.

5. A low-frequency noise blind pixel detection system based on segmented noise difference, employing the low-frequency noise blind pixel detection method based on segmented noise difference as described in any one of claims 1-4, characterized in that, include: The data acquisition module is used to observe all pixels of the detector against a uniform and stable blackbody to obtain observation data for each pixel. f represents the frame number; The preliminary calculation module is used to calculate the observation data for each pixel. noise ; The recombination module is used to reconstruct the observation data. After segmenting, the data is combined, and the noise of each new combined data segment is calculated. ; The index calculation module is used to calculate the index based on the observation data of each pixel. noise and the noise in each new data segment after combination Calculate the first low-frequency stability index for each pixel. Second low-frequency stability index ; The blind cell determination module is used to determine the first low-frequency stability index. Below the first low-frequency stability index threshold and the second low-frequency stability index Pixels below the second low-frequency stability index threshold are classified as low-frequency noise blind pixels.

6. An electronic device, characterized in that, include: The memory is used to store the processing program; A processor, which, when executing the processing program, implements the low-frequency noise blind cell detection method based on segmented noise differences as described in any one of claims 1 to 4.

7. A readable storage medium, characterized in that, The readable storage medium stores a processing program, which, when executed by a processor, implements the low-frequency noise blind cell detection method based on segmented noise differences as described in any one of claims 1 to 4.

Images