Unmanned aerial vehicle CVBS video signal efficient detection method and device

Abstract

This invention discloses an efficient method and apparatus for detecting CVBS video signals from unmanned aerial vehicles (UAVs). The method includes: acquiring spatial electromagnetic signals; the spatial electromagnetic signals being spatial electromagnetic radiation signals to be detected as containing simulated FPV UAV image transmission signals; classifying and extracting the spatial electromagnetic signals to obtain a set of branched signals; the set of branched signals including a luminance signal to be detected, a chrominance signal to be detected, and a synchronization signal to be detected; detecting the set of branched signals to obtain detection result information; the detection result information being used to characterize whether the spatial electromagnetic signals contain simulated FPV UAV image transmission signals.

Description

Technical Field

[0001] This invention relates to the fields of intelligent signal processing, strategy optimization, and signal recognition, specifically to a method and apparatus for efficient detection of CVBS video signals from unmanned aerial vehicles (UAVs). Background Technology

[0002] FPV drones are drone systems that enable operators to control the drone with low latency and precision from a first-person perspective using real-time video transmission and remote control technology. They use cameras to capture images and convert them into analog transmission signals, which are then transmitted via radio waves. The receiving end demodulates and reconstructs the analog transmission signals, ultimately displaying them as images on the remote control.

[0003] The signal transmission link of an FPV drone's data link is divided into a downlink and an uplink. The uplink refers to sending remote control commands from the drone's control console to the drone. The downlink is the process of the drone sending relevant information and data to its control console. FPV drones primarily use the downlink transmission method because it employs a unidirectional transmission approach, allowing for real-time image transmission without establishing a connection. This method results in lower latency and higher real-time performance for FPV image transmission.

[0004] The simulated image transmission signal of an FPV drone essentially employs CVBS signal modulation with a carrier frequency of 5.8 GHz. The broadband characteristics of the carrier and the complexity of the modulation present numerous challenges for traditional signal detection and identification methods. How to quickly and accurately detect CVBS signals based on the characteristics of the simulated image transmission signal of an FPV drone is a problem that needs to be solved. Summary of the Invention

[0005] This invention primarily addresses the problem of how to quickly and accurately detect CVBS signals based on the simulated image transmission signal characteristics of FPV UAVs. This invention discloses an efficient method and apparatus for detecting UAV CVBS video signals.

[0006] In a first aspect, this invention discloses a method for efficient detection of CVBS video signals from unmanned aerial vehicles (UAVs), comprising:

[0007] S1, acquire spatial electromagnetic signals; the spatial electromagnetic signals are spatial electromagnetic radiation signals that are to be detected to determine whether they contain simulated image transmission signals of FPV UAVs.

[0008] S2, classify and extract the spatial electromagnetic signals to obtain a set of branched signals; the set of branched signals includes the luminance signal to be detected, the chrominance signal to be detected, and the synchronization signal to be detected;

[0009] S3, the branch signal set is detected to obtain detection result information; the detection result information is used to characterize whether the spatial electromagnetic signal contains FPV UAV simulated image transmission signal.

[0010] The process of classifying and extracting the spatial electromagnetic signals to obtain a set of branched signals includes:

[0011] S21, Perform synchronization signal extraction processing on the spatial electromagnetic signal to obtain the synchronization signal to be detected;

[0012] S22, based on the synchronization signal to be detected and the spatial electromagnetic signal, perform brightness signal extraction processing to obtain the brightness signal to be detected;

[0013] S23, perform chromaticity signal extraction processing on the spatial electromagnetic signal to obtain the chromaticity signal to be detected;

[0014] S24, using the detection synchronization signal, detection luminance signal and detection chrominance signal, a set of branch signals is constructed.

[0015] The detection of the branched signal set to obtain detection result information includes:

[0016] S31, perform image representation processing on the split signal set to obtain an image sequence set;

[0017] S32, using a preset neural network model, perform detection processing on the image sequence set to obtain a first detection result set;

[0018] S33, using a preset image template, perform detection processing on the image sequence set to obtain a second detection result set;

[0019] S34, perform a fusion calculation on the first detection result set and the second detection result set to obtain the detection result information.

[0020] The image representation processing of the split signal set to obtain an image sequence set includes:

[0021] Set time interval information;

[0022] Using the time interval information, the set of branched signals is subjected to time-domain segmentation to obtain a branched signal subset corresponding to each time interval; the branched signal subset includes a detection synchronization sub-signal, a detection luminance sub-signal, and a detection chrominance sub-signal obtained by performing time-domain segmentation on the detection synchronization signal, the detection luminance signal, and the detection chrominance signal, respectively.

[0023] Each branch signal subset is processed to obtain a corresponding image;

[0024] For the images corresponding to all branched signal subsets, an image sequence set is constructed.

[0025] The step of using a preset image template to perform detection processing on the image sequence set to obtain a second detection result set includes:

[0026] S331, using a preset image template, perform pixel value distribution detection processing on each image in the image sequence set to obtain the corresponding first detection result value;

[0027] S332, using a preset image template, perform linear distribution detection processing on each image in the image sequence set to obtain the corresponding second detection result value;

[0028] S333, perform a first fusion calculation on the first detection result value and the second detection result value of all images to obtain a second detection result set.

[0029] The step of using a preset image template to perform pixel value distribution detection processing on each image in the image sequence set to obtain the corresponding first detection result value includes:

[0030] S3311, For each channel of the preset image template, perform value distribution statistical processing to obtain the corresponding standard value distribution information; the value distribution statistical processing is to statistically obtain the distribution probability of pixel points of the image channel in each value interval.

[0031] S3312, For each channel of each image in the image sequence set, perform value distribution statistical processing to obtain the corresponding value distribution information;

[0032] S3313, perform difference calculation on the value distribution information of all channels of each image in the image sequence set to obtain the difference value of the image. cp and deviation value xp ;

[0033] The difference calculation includes:

[0034] ,

[0035] ,

[0036] in, and Let be the probability distribution of the j-th channel of the image and the image template in the i-th value interval, respectively. Let be the mean of the differences between the probability distributions of the images and image templates in the j-th channel of the image sequence set. Let M be the variance of the difference between the probability distributions of the images and image templates in the j-th channel of the image sequence set, where M is the total number of channels and N is the total number of value intervals.

[0037] S3314, using the difference value and deviation value of the image, construct the first detection result value of the image.

[0038] The step of using a preset image template to perform linear distribution detection processing on each image in the image sequence set to obtain the corresponding second detection result value includes:

[0039] S3321, For the preset image template and the images in the image sequence set, feature lines are extracted in each channel to obtain the corresponding first feature line and second feature line;

[0040] S3322, Local feature points are extracted from the first feature line and the second feature line respectively to obtain the corresponding first feature point and second feature point;

[0041] S3323 calculates the anomaly difference for the feature lines and feature points of all channels of an image, and obtains the corresponding graphic difference value and graphic deviation value.

[0042] S3324, using the image difference value and image deviation value of the image, a corresponding second detection result value is constructed.

[0043] A second aspect of this invention discloses a high-efficiency detection device for CVBS video signals from unmanned aerial vehicles (UAVs), the device comprising:

[0044] Memory containing executable program code;

[0045] A processor coupled to the memory;

[0046] The processor calls the executable program code stored in the memory to execute the efficient detection method for UAV CVBS video signals.

[0047] In a third aspect of this invention, a computer-storable medium is disclosed, wherein the computer-storable medium stores computer instructions, which, when invoked by a computer, are used to execute the aforementioned efficient detection method for UAV CVBS video signals.

[0048] In a fourth aspect of this invention, an information data processing terminal is disclosed, which is used to implement the aforementioned method for efficient detection of UAV CVBS video signals.

[0049] The beneficial effects of this invention are as follows:

[0050] The present invention converts a subset of time-domain split signals into an image sequence (synchronization signal) using Gram angle field transform. R channel, brightness G channel, chroma The B-channel enables the spatial representation of time-domain signals. The advantages of this conversion are: it maps the time-domain features of the signal (such as pulse timing and frequency variations) to the spatial features of the image (pixel distribution and channel correlation), adapting to both neural networks (excelling in visual tasks) and template matching (image feature comparison) detection paradigms; it preserves the continuity of the signal in the time dimension (through time interval segmentation and sequence construction), capturing the dynamic characteristics of image transmission signals (such as the temporal patterns of image changes), and improving the detection capability of dynamic signals.

[0051] This invention improves the robustness and accuracy of signal detection through multimodal detection fusion. The innovative approach combines neural network detection and image template detection methods, outputting the final result through fusion computation, thus avoiding the limitations of a single detection method: the neural network (ResNet) is trained on a large set of standard images, excelling at learning complex features and generalizing to unknown scenarios, and can handle distortions caused by signal interference; image template detection, based on image templates of standard signals, excels at capturing standard signal features through precise matching of pixel distribution and linear features, exhibiting high stability; the fusion computation simultaneously considers the detection results and confidence levels of both methods, highlighting the contribution of high-confidence results through a weighted strategy, reducing the risk of false positives / false negatives. Attached Figure Description

[0052] Figure 1 This is a flowchart illustrating the implementation of the method of the present invention. Detailed Implementation

[0053] To better understand the content of this invention, an embodiment is provided here.

[0054] Figure 1 This is a flowchart illustrating the implementation of the method of the present invention.

[0055] In a first aspect, this invention discloses a method for efficient detection of CVBS video signals from unmanned aerial vehicles (UAVs), comprising:

[0056] S1, acquire spatial electromagnetic signals; the spatial electromagnetic signals are spatial electromagnetic radiation signals that are to be detected to determine whether they contain simulated image transmission signals of FPV UAVs.

[0057] S2, classify and extract the spatial electromagnetic signals to obtain a set of branched signals; the set of branched signals includes the luminance signal to be detected, the chrominance signal to be detected, and the synchronization signal to be detected;

[0058] S3, the branch signal set is detected to obtain detection result information; the detection result information is used to characterize whether the spatial electromagnetic signal contains FPV UAV simulated image transmission signal;

[0059] The process of classifying and extracting the spatial electromagnetic signals to obtain a set of branched signals includes:

[0060] S21, Perform synchronization signal extraction processing on the spatial electromagnetic signal to obtain the synchronization signal to be detected;

[0061] S22, based on the synchronization signal to be detected and the spatial electromagnetic signal, perform brightness signal extraction processing to obtain the brightness signal to be detected;

[0062] S23, perform chromaticity signal extraction processing on the spatial electromagnetic signal to obtain the chromaticity signal to be detected;

[0063] S24, using the detection synchronization signal, detection luminance signal and detection chrominance signal, a set of branch signals is constructed;

[0064] The process of extracting synchronization signals from the spatial electromagnetic signals to obtain the synchronization signal to be detected includes: the amplitude of the synchronization signal is usually much higher than that of the luminance and chrominance signals (for example, in a standard CVBS signal, the synchronization pulse is -0.3V, while the image signal range is 0~0.7V). By setting a voltage threshold circuit, when the signal is lower than the threshold, it is determined to be a synchronization pulse, and the synchronization signal can be separated.

[0065] S22 includes: "cutting off" the synchronization pulse through a clamping circuit, retaining the 0~0.7V image signal portion; then filtering out the chroma signal: in the NTSC system, the chroma signal (C) is modulated and superimposed on the luminance signal with a 3.58MHz subcarrier; in the PAL system, the chroma signal is modulated with a 4.43MHz subcarrier. A low-pass filter (cutoff frequency approximately 3MHz) is used to filter out the high-frequency chroma signal, retaining the low-frequency luminance signal (Y).

[0066] S23 includes: using a bandpass filter with a center frequency corresponding to the subcarrier (NTSC 3.58MHz, PAL 4.43MHz) to extract high-frequency chroma signals from the image signal after removing the synchronization signal (filtering out low-frequency luminance signals), or the extracted chroma signal is a modulated subcarrier signal, which needs to be demodulated by a synchronization detection circuit (using the color synchronization signal to recover the subcarrier phase) to obtain two color difference signals (such as the I and Q signals of NTSC, and the U and V signals of PAL), and finally obtain complete chroma information.

[0067] The detection of the branched signal set to obtain detection result information includes:

[0068] S31, perform image representation processing on the split signal set to obtain an image sequence set;

[0069] S32, using a preset neural network model, perform detection processing on the image sequence set to obtain a first detection result set;

[0070] S33, using a preset image template, perform detection processing on the image sequence set to obtain a second detection result set;

[0071] S34, perform a fusion calculation on the first detection result set and the second detection result set to obtain the detection result information.

[0072] The step of fusing the first detection result set and the second detection result set to obtain detection result information includes:

[0073] S341, Perform detection fusion calculation on the first detection result set and the second detection result set of each image to obtain the corresponding fused detection value;

[0074] Determine whether the fusion detection value is greater than a preset discrimination threshold. If it is greater, determine that there is an FPV drone simulated image transmission signal within the acquisition time of the branch signal subset corresponding to the image of the fusion detection value. If it is not greater, determine that there is no FPV drone simulated image transmission signal within the acquisition time of the branch signal subset corresponding to the image of the fusion detection value.

[0075] S342, using the fused detection values ​​of all images, constructs the detection result information.

[0076] The expression for the detection fusion calculation is:

[0077] ,

[0078] in, Let i be the fusion detection value corresponding to the i-th image. and These represent the first detection result and the corresponding confidence value, respectively.

[0079] This formula uses a logarithmic function to perform a non-linear mapping of confidence levels, achieving the effect of "amplifying the weight of high-confidence results and suppressing the weight of low-confidence results." When the confidence level is close to 1 (high confidence), or A small absolute value but a positive sign can effectively enhance the corresponding detection results. or The contribution; when the confidence level is low, Negative values ​​weaken the interference of low-confidence results and prevent a single unreliable result from affecting the final judgment. By linearly superimposing the weighted results of the two detection methods, the effective fusion of multi-source information is achieved, improving the overall robustness of the detection.

[0080] The image representation processing of the split signal set to obtain an image sequence set includes:

[0081] Set time interval information;

[0082] Using the time interval information, the set of branched signals is subjected to time-domain segmentation to obtain a branched signal subset corresponding to each time interval; the branched signal subset includes a synchronization sub-signal to be detected, a luminance sub-signal to be detected, and a chrominance sub-signal to be detected obtained by time-domain segmentation.

[0083] Each branch signal subset is processed to obtain a corresponding image;

[0084] The images corresponding to all the branch signal subsets are fused to obtain a set of image sequences.

[0085] The image representation processing involves performing Gram angle field transformation on the acquisition time information of the branched signal subsets to obtain the transformation result, and then rounding and normalizing the transformation result to obtain a two-dimensional coordinate matrix of the image. Each acquisition time corresponds to a two-dimensional coordinate in the two-dimensional coordinate matrix. The detection sub-signal, detection luminance sub-signal, and detection chrominance sub-signal in the branched signal subset are represented as the R channel value, G channel value, and B channel value of the image, respectively. According to the acquisition time of each sub-signal, the channel value corresponding to the acquisition time is set on the corresponding two-dimensional coordinate at that time to complete the image representation and obtain the corresponding image.

[0086] The rounding can be either rounding up or rounding down;

[0087] The preset neural network model is a ResNet network trained using a standard image set.

[0088] The step of using a preset neural network model to perform detection processing on the image sequence set to obtain a first detection result set involves inputting each image in the image sequence set into the neural network model, outputting the first detection result and related confidence value for each image, and constructing the first detection result set using the first detection results and related confidence values ​​of all images.

[0089] The standard image set is an image set obtained by performing image representation processing using simulated image transmission signals from a standard FPV UAV.

[0090] The image template is an image obtained by simulating image transmission signals from a standard FPV drone and performing image representation processing.

[0091] The step of using a preset image template to perform detection processing on the image sequence set to obtain a second detection result set includes:

[0092] S331, using a preset image template, perform pixel value distribution detection processing on each image in the image sequence set to obtain the corresponding first detection result value;

[0093] S332, using a preset image template, perform linear distribution detection processing on each image in the image sequence set to obtain the corresponding second detection result value;

[0094] S333, perform a first fusion calculation on the first detection result value and the second detection result value of all images to obtain a second detection result set.

[0095] The step of using a preset image template to perform pixel value distribution detection processing on each image in the image sequence set to obtain the corresponding first detection result value includes:

[0096] S3311, For each channel of the preset image template, perform value distribution statistical processing to obtain the corresponding standard value distribution information; the value distribution statistical processing is to statistically obtain the distribution probability of pixel points of the image channel in each value interval.

[0097] S3312, For each channel of each image in the image sequence set, perform value distribution statistical processing to obtain the corresponding value distribution information;

[0098] S3313, perform difference calculation on the value distribution information of all channels of each image in the image sequence set to obtain the difference value of the image. cp and deviation value xp ;

[0099] The channel, or image channel, refers to the R channel image, G channel image, and B channel image of an image;

[0100] The difference calculation includes:

[0101] ,

[0102] ,

[0103] in, and Let be the probability distribution of the j-th channel of the image and the image template in the i-th value interval, respectively. Let be the mean of the differences between the probability distributions of the images and image templates in the j-th channel of the image sequence set. Let M be the variance of the difference between the probability distributions of the images and image templates in the j-th channel of the image sequence set, where M is the total number of channels and N is the total number of value intervals.

[0104] The difference calculation expression, through a combination of "absolute value + sine function," accurately quantifies the difference in pixel value distribution between the image to be detected and the template, while enhancing the sensitivity to significant differences. The absolute value term directly measures the difference in probability distribution within a single interval; the sine function term changes approximately linearly when the difference is small, avoiding excessive amplification of minor noise; when the difference is large, the sine function value tends to stabilize (maximum value 1), highlighting significant differences while avoiding saturation interference from extreme values ​​on the overall calculation; the double summation covers all channels and value intervals, comprehensively reflecting the overall difference in pixel distribution, providing accurate quantitative basis for template matching.

[0105] The expression for the graphic difference value measures the overall offset of the pixel distribution using the formula "mean - variance ratio + logarithmic function," taking into account both the central tendency and dispersion of the distribution. The mean difference reflects the overall direction of the distribution offset, while the variance reflects the dispersion of the offset. The "significance" of normalizable bias (the smaller the variance, the more significant the bias for the same mean); the logarithmic function non-linearly adjusts the bias ratio, when... When the ratio is greater than 1 (significant offset), the logarithmic output is positive and increases with the ratio, amplifying the effect of significant offset; when the ratio is close to 1 (small offset), the logarithmic output is close to 0, reducing interference; the summation covers all channels, comprehensively reflecting the overall deviation of the multi-channel distribution, providing a global quantitative indicator for judging whether the signal meets the standard characteristics.

[0106] S3314, using the difference value and deviation value of the image, construct the first detection result value of the image.

[0107] In S3311 and S3312, the value range used for the statistical processing of value distribution is the same.

[0108] The image template includes channel images of the G channel, R channel, and B channel;

[0109] The step of using a preset image template to perform linear distribution detection processing on each image in the image sequence set to obtain the corresponding second detection result value includes:

[0110] S3321, For the preset image template and the images in the image sequence set, feature lines are extracted in each channel to obtain the corresponding first feature line and second feature line;

[0111] S3322, Local feature points are extracted from the first feature line and the second feature line respectively to obtain the corresponding first feature point and second feature point;

[0112] S3323 calculates the anomaly difference for the feature lines and feature points of all channels of an image, and obtains the corresponding graphic difference value and graphic deviation value.

[0113] S3324, using the image difference value and image deviation value of the image, a corresponding second detection result value is constructed.

[0114] The expression for calculating the abnormal differences is:

[0115] ,

[0116]

[0117] in, yc For the difference in the graph, tc This represents the deviation value of the graph. This is the difference vector between the direction vectors of the first feature line of the j-th channel of an image and the second feature line of the j-th channel of a preset image template. This represents the difference vector between the coordinate vectors of the first feature point in the j-th channel of an image and the second feature point in the j-th channel of a preset image template. Represents a vector Find the modulus. This indicates the cross product operation between two vectors. This indicates finding the cosine value of two vectors.

[0118] In the graphical difference values, For vectors with vector The cosine value of the vector is in the range of [-1, 1]. The closer the value is to 1, the more consistent the direction is (the smaller the difference). The denominator normalizes the vector magnitude to eliminate the interference of the direction similarity caused by the difference in feature scale (such as the change in vector magnitude caused by different signal strength). The summation covers all channels and integrates the matching degree of multi-channel structural features to effectively capture stable line / edge features (such as the periodic lines of the synchronization pulse) in the analog image transmission signal.

[0119] The calculation of the graph deviation value quantifies the degree of vertical deviation of the feature by the magnitude ratio of the vector cross product, complementing the graph difference value (directional consistency) to comprehensively characterize the differences in structural features. The magnitude of the vector cross product ranges from [0,1], with a larger value indicating a higher degree of perpendicularity between the two vectors (more significant deviation); it complements the graph difference value in terms of angle, jointly and completely describing the spatial relationship of the feature vectors, avoiding the one-sidedness of a single angle indicator; by integrating multi-channel results, it accurately captures signal structural distortions (such as line breaks caused by the loss of synchronization signals).

[0120] The first fusion computing process includes:

[0121] ,

[0122] ,

[0123] in, The second detection result is for the i-th image in the image sequence set. Let be the confidence value of the i-th image in the image sequence set. and These are the difference value and deviation value of the i-th image, respectively. and These are the graphic difference value and graphic deviation value of the i-th image, respectively; the second detection result set includes the second detection result and confidence value of each image.

[0124] The feature line extraction can be performed using the probabilistic Hough transform method, feature point clustering and line fitting method;

[0125] The local feature point extraction can be performed using gradient peak extraction or SIFT / ORB point extraction.

[0126] A second aspect of this invention discloses a high-efficiency detection device for CVBS video signals from unmanned aerial vehicles (UAVs), the device comprising:

[0127] Memory containing executable program code;

[0128] A processor coupled to the memory;

[0129] The processor calls the executable program code stored in the memory to execute the efficient detection method for UAV CVBS video signals.

[0130] In a third aspect of this invention, a computer-storable medium is disclosed, wherein the computer-storable medium stores computer instructions, which, when invoked by a computer, are used to execute the aforementioned efficient detection method for UAV CVBS video signals.

[0131] In a fourth aspect of this invention, an information data processing terminal is disclosed, which is used to implement the aforementioned method for efficient detection of UAV CVBS video signals.

[0132] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A method for efficient detection of CVBS video signals from unmanned aerial vehicles (UAVs), characterized in that, include: S1, acquire spatial electromagnetic signals; the spatial electromagnetic signals are spatial electromagnetic radiation signals that are to be detected to determine whether they contain simulated image transmission signals of FPV UAVs. S2, classify and extract the spatial electromagnetic signals to obtain a set of branched signals; the set of branched signals includes the luminance signal to be detected, the chrominance signal to be detected, and the synchronization signal to be detected; S3, Detect the set of branched signals to obtain detection result information; The detection result information is used to characterize whether the spatial electromagnetic signal contains FPV drone simulated image transmission signals, including: S31, perform image representation processing on the split signal set to obtain an image sequence set, including: Set time interval information; Using the time interval information, the set of branched signals is subjected to time-domain segmentation to obtain a branched signal subset corresponding to each time interval; the branched signal subset includes a detection synchronization sub-signal, a detection luminance sub-signal, and a detection chrominance sub-signal obtained by performing time-domain segmentation on the detection synchronization signal, the detection luminance signal, and the detection chrominance signal, respectively. Image representation processing is performed on each subset of branched signals to obtain the corresponding image. This image representation processing involves performing Gram angle field transformation on the acquisition time information of the branched signal subsets to obtain the transformation result, and then rounding and normalizing the transformation result to obtain a two-dimensional coordinate matrix of the image. Each acquisition time corresponds to a two-dimensional coordinate in the two-dimensional coordinate matrix. The detection sub-signal, detection luminance sub-signal, and detection chrominance sub-signal in the branched signal subsets are represented as the R channel value, G channel value, and B channel value of the image, respectively. Based on the acquisition time of each sub-signal, the channel value corresponding to that acquisition time is set on the corresponding two-dimensional coordinate at that time to complete the image representation and obtain the corresponding image. For each subset of signals corresponding to a given image, a set of image sequences is constructed. S32, using a preset neural network model, perform detection processing on the image sequence set to obtain a first detection result set; S33, using a preset image template, perform detection processing on the image sequence set to obtain a second detection result set; S34, perform a fusion calculation on the first detection result set and the second detection result set to obtain the detection result information.

2. The efficient detection method for UAV CVBS video signals as described in claim 1, characterized in that, The process of classifying and extracting the spatial electromagnetic signals to obtain a set of branched signals includes: S21, Perform synchronization signal extraction processing on the spatial electromagnetic signal to obtain the synchronization signal to be detected; S22, based on the synchronization signal to be detected and the spatial electromagnetic signal, perform brightness signal extraction processing to obtain the brightness signal to be detected; S23, perform chromaticity signal extraction processing on the spatial electromagnetic signal to obtain the chromaticity signal to be detected; S24, using the detection synchronization signal, detection luminance signal and detection chrominance signal, a set of branch signals is constructed.

3. The efficient detection method for UAV CVBS video signals as described in claim 1, characterized in that, The step of using a preset image template to perform detection processing on the image sequence set to obtain a second detection result set includes: S331, using a preset image template, perform pixel value distribution detection processing on each image in the image sequence set to obtain the corresponding first detection result value; S332, using a preset image template, perform linear distribution detection processing on each image in the image sequence set to obtain the corresponding second detection result value; S333, perform a first fusion calculation on the first detection result value and the second detection result value of all images to obtain a second detection result set.

4. The efficient detection method for UAV CVBS video signals as described in claim 3, characterized in that, The step of using a preset image template to perform pixel value distribution detection processing on each image in the image sequence set to obtain the corresponding first detection result value includes: S3311, For each channel of the preset image template, perform value distribution statistical processing to obtain the corresponding standard value distribution information; the value distribution statistical processing is to statistically obtain the distribution probability of pixel points of the image channel in each value interval. S3312, For each channel of each image in the image sequence set, perform value distribution statistical processing to obtain the corresponding value distribution information; S3313, perform difference calculation on the value distribution information of all channels of each image in the image sequence set to obtain the difference value of the image. cp and deviation value xp ; The difference calculation includes: , , in, and Let be the probability distribution of the j-th channel of the image and the image template in the i-th value interval, respectively. Let be the mean of the differences between the probability distributions of the images and image templates in the j-th channel of the image sequence set. Let M be the variance of the difference between the probability distributions of the images and image templates in the j-th channel of the image sequence set, where M is the total number of channels and N is the total number of value intervals. S3314, using the difference value and deviation value of the image, construct the first detection result value of the image.

5. The efficient detection method for UAV CVBS video signals as described in claim 3, characterized in that, The step of using a preset image template to perform linear distribution detection processing on each image in the image sequence set to obtain the corresponding second detection result value includes: S3321, For the preset image template and the images in the image sequence set, feature lines are extracted in each channel to obtain the corresponding first feature line and second feature line; S3322, Local feature points are extracted from the first feature line and the second feature line respectively to obtain the corresponding first feature point and second feature point; S3323 calculates the anomaly difference for the feature lines and feature points of all channels of an image, and obtains the corresponding graphic difference value and graphic deviation value. S3324, using the image difference value and image deviation value of the image, a corresponding second detection result value is constructed.

6. A high-efficiency detection device for CVBS video signals from unmanned aerial vehicles (UAVs), characterized in that, The device includes: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the UAV CVBS video signal high-efficiency detection method as described in any one of claims 1 to 5.

7. A computer-storable medium, characterized in that, The computer storage medium stores computer instructions, which, when invoked by the computer, are used to execute the UAV CVBS video signal high-efficiency detection method as described in any one of claims 1 to 5.

8. An information data processing terminal, characterized in that, The information data processing terminal is used to implement the efficient detection method for UAV CVBS video signals as described in any one of claims 1 to 5.

Images