A stable precision testing method, system, medium, device and terminal

Abstract

The present invention belongs to the technical field of stability accuracy testing, and discloses a stability accuracy testing method, system, medium, device and terminal. By collecting the original video image data of an optoelectronic camera and processing the image to obtain a template; after respectively segmenting and binarizing the image, the template matching method is used to identify the video, and the feature data of each frame of image in the video is obtained, and then the stability accuracy data in the entire video is calculated using a formula. The present invention proposes a complete method of image recognition in stability accuracy testing, which solves the problem of incomplete image processing in actual stability accuracy testing. At the same time, the present invention first uses the template matching method (matchTemplate function) to identify the effective features of the image in stability accuracy processing, and the accuracy is accurate to the pixel level and accurate. The accuracy used by others for image recognition is sub-pixel level.

Description

Technical Field

[0001] The present invention belongs to the technical field of stability accuracy testing, and particularly relates to a stability accuracy testing method, system, medium, device and terminal. Background Art

[0002] Currently, the existing method for detecting stability accuracy is to set up a target in the room, use a pod or a pan-tilt to carry an external imaging device to photograph the target, and calculate the stability accuracy by using the distance between the center of the image and the center of the target position. This is a post-processing method (post-processing after shooting), lacking real-time performance, having a long test duration, low efficiency, and low measurement accuracy.

[0003] Chinese Patent CN201610554789 discloses a method and system for detecting the stability accuracy of a pod. The carrier of the pod is fixed by a vibration table, and a vibration excitation is provided for the carrier of the pod; an optical signal is emitted to the pod, and the vibration amplitude of the pod is detected by detecting the reflected optical signal, and a stability accuracy indication parameter is output according to the detection result. This patent can only measure the vibration in the coaxial direction (front and back) of the emitted light by using the reflected light, and cannot measure the vibration in the vertical direction (left, right, up and down), and the measurement data is not comprehensive enough. Affected by the surface material, shape, etc. of the item to be measured, there will be a large deviation between the measured data and the actual value.

[0004] Chinese Patent CN200710098921.6 discloses a precision detector, which mainly includes a locator, a positioning plate, a tetrahedron, a flat plate, an angle plate and a mandrel. The locator is placed on the platform to make the locator in a horizontal state; the flat plate is placed on the locator and positioned by the locator; the mandrel is installed on the angle plate, and the tetrahedron is pressed on the mandrel through a fastener; the angle plate is installed on the flat plate positioned by the locator, and the angle plate is positioned by using a guide ruler and the locator and then fixed on the flat plate; the components assembled by the flat plate, the tetrahedron, the mandrel and the angle plate are installed on the instrument installation base surface of the inner frame of the turntable to be measured, and the position of the flat plate on the instrument installation base surface is adjusted by the positioning plate and then the flat plate is fixed, so as to detect the position accuracy of the turntable. The present invention conducts on-site calibration and detection of the turntable accuracy of the unit test system, is simple and convenient to use, does not require special detection personnel, and general technical personnel can use it after a short training, improving the detection efficiency and accuracy. This kind of detector determines the accuracy through instrument detection, and there are errors in this detection method, the accuracy is not accurate and the detected data is not comprehensive enough. Therefore, it is urgent to design a new stability accuracy testing method and system.

[0005] Through the above analysis, the problems and defects existing in the prior art are as follows:

[0006] (1) The existing method for detecting stability accuracy is a post-processing method (post-processing after shooting), lacking real-time performance, having a long test duration, low efficiency, and low measurement accuracy.

[0007] (2) The prior art can only measure the vibration in the coaxial direction of the emitted light by using reflected light measurement, and cannot measure the vibration in the vertical direction, so the measurement data is not comprehensive enough; affected by the surface material, shape, etc. of the item to be measured, there will be a large deviation between the measured data and the actual value.

[0008] (3) The existing detector determines the accuracy through instrument detection. This detection method has errors, the accuracy is inaccurate, and the detected data is not comprehensive enough. Summary of the Invention

[0009] In view of the problems existing in the prior art, the present invention provides a stable accuracy test method, system, medium, device and terminal, and particularly relates to a stable accuracy test method, system, medium, device and terminal based on image recognition.

[0010] The present invention is implemented as follows. A stable accuracy test method, the stable accuracy test method includes:

[0011] Collect the original video image data of the optoelectronic camera, and process the image to obtain a template; after segmenting and binarizing the image respectively, use the template matching method to identify the video, obtain the feature data of each frame of the image in the video, and then use the formula to calculate the stable accuracy data in the entire video.

[0012] Further, the stable accuracy test method further includes:

[0013] The optoelectronic camera on the stable platform tracks the relatively moving target target, and the original video output by the collected optoelectronic camera is divided into the following two paths for subsequent work:

[0014] (1) Capture a single-frame image in the video. By segmenting and binarizing the image, and combining the use of the THRESH_OTSU and THRESH_BINARY functions, a series of target templates are obtained to form a target template library; compare the images in the target template library, and select the optimal target template as the template for template matching.

[0015] (2) Select the video segment to be calculated according to the scheme, and divide the video segment into single-frame images.

[0016] Further, the stable accuracy test method further includes:

[0017] (1) Use the template matching method, utilize the matchTemplate function, by sliding the template image on each frame of the target image one by one for comparison, use the TM_SQDIFF_NORMED function for image matching, find the most suitable matching part, and perform normalization processing;

[0018] (2) Use the minMaxLoc function to find the matching results and positions of the maximum and minimum values in the matrix, and obtain the feature data or located coordinates recognized by the template in a single-frame image; substitute the known field of view angle of the optoelectronic camera into the formula for calculation to obtain the stable accuracy data for an entire segment.

[0019] Further, the stable accuracy test method includes the following steps:

[0020] Step 1, collect the original video data output by the optoelectronic camera on the stable platform when tracking the target target.

[0021] Step 2, capture a series of single-frame images, and perform segmentation and binarization processing on the images respectively to obtain a target template library; at the same time, intercept an appropriate length of the original video segment, and process the video by removing background noise and local magnification.

[0022] Step 3, select the target template with the best effect as the template, and at the same time extract each frame of the video image, and perform segmentation and binarization processing on the images respectively.

[0023] Step 4, use the template matching method to match the template image block with the input image, find the most suitable matching part and perform normalization processing, and analyze and calculate to obtain the feature data of the image.

[0024] Step 5, substitute the field of view angle, and calculate the stable accuracy value of the entire video according to the formula.

[0025] Further, the segmentation and binarization processing of the images in step 2 includes:

[0026] (1) Receive a series of single-frame images captured, obtain the class labels and binarization segmentation labels of the single-frame images, and perform feature extraction on the single-frame images to obtain a training data set.

[0027] (2) Obtain a target heat map through a preset deep learning convolutional neural network, and obtain an image segmentation loss value according to the target heat map and the binarization segmentation label.

[0028] (3) Adjust the parameters of the preset deep learning convolutional neural network according to the training data set and the image segmentation loss value, and then complete the training of the model.

[0029] (4) Use the trained deep learning convolutional neural network model to perform segmentation and binarization processing on a series of single-frame images captured respectively, and output the processing results.

[0030] Further, the substituting the field of view angle and calculating the stable accuracy value of the entire video in step 5 includes:

[0031] The displacement deviation of the crosshair center between two adjacent frames in the video image is denoted as Δx i and Δy i . Calculate 1 stable accuracy value per second and statistically average it over 1 minute;

[0032] The stable accuracy per second is calculated according to the following formula:

[0033]

[0034]

[0035]

[0036] where, Δx i represents the number of pixel deviations in the azimuth of the crosshair center between two frames of images; Δy i represents the number of pixel deviations in the pitch of the crosshair center between two frames of images; θ x represents the measured azimuth field of view angle of the small field of view, with the unit of °; θ y represents the measured pitch and roll field of view angle of the minimum field of view, with the unit of °; X represents the resolution in the azimuth direction of the field of view; Y represents the resolution in the pitch direction of the field of view; μ xt represents the azimuth stability accuracy, with the unit of mrad; μ yt represents the pitch stability accuracy, with the unit of mrad; μ t represents the stability accuracy per second, with the unit of mrad; n represents the total number of frames;

[0037] The stability accuracy is:

[0038]

[0039] Another object of the present invention is to provide a stability accuracy test system applying the described stability accuracy test method. The stability accuracy test system includes:

[0040] An original video acquisition module for acquiring the original video data output by an optoelectronic camera on a stable platform when tracking a target;

[0041] A target template library construction module for capturing a series of single-frame images and respectively performing segmentation and binarization processing on the images to obtain a target template library;

[0042] A video processing module for intercepting an original video segment of an appropriate length and processing the video by removing background noise and local magnification;

[0043] A template selection module for selecting the best target template as the template;

[0044] An image processing module for extracting each frame of the video and respectively performing segmentation and binarization processing on the images;

[0045] An image feature data calculation module, which is used to use the template matching method to match the template image block with the input image, find the most suitable matching part and perform normalization processing, and analyze and calculate the feature data of the image;

[0046] A stable accuracy numerical calculation module, which is used to substitute the field of view angle and calculate the stable accuracy value of the entire video according to the formula.

[0047] Another object of the present invention is to provide a computer device, which includes a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor executes the following steps:

[0048] By collecting the original video image data of the optoelectronic camera and processing the image to obtain a template; after the image is segmented and binarized respectively, the template matching method is used to identify the video, obtain the feature data of each frame of the image in the video, and then use the formula to calculate the stable accuracy data in the entire video.

[0049] Another object of the present invention is to provide a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the following steps:

[0050] By collecting the original video image data of the optoelectronic camera and processing the image to obtain a template; after the image is segmented and binarized respectively, the template matching method is used to identify the video, obtain the feature data of each frame of the image in the video, and then use the formula to calculate the stable accuracy data in the entire video.

[0051] Another object of the present invention is to provide an information data processing terminal, which is used to implement the stable accuracy test system described above.

[0052] Combined with the above technical solutions and the technical problems solved, please analyze the advantages and positive effects of the technical solution to be protected by the present invention from the following aspects:

[0053] First, aiming at the technical problems existing in the above-mentioned prior art and the difficulty of solving this problem, closely combining the technical solution to be protected by the present invention and the results and data in the R & D process, etc., analyze in detail and profoundly how the technical solution of the present invention solves the technical problems and the creative technical effects brought after solving the problems. The specific description is as follows:

[0054] The present invention collects the original video of an optoelectronic camera, processes the image to obtain a template, uses the template matching method to identify the video, obtains the feature data of each frame of the image in the video, and then calculates the stability accuracy data of the entire video using a formula. In the stability accuracy processing, the present invention first uses the template matching method (matchTemplate function) to identify the effective features of the image, with an accurate pixel-level precision and accuracy. The accuracy used by others for image recognition is sub-pixel level, and this level of data is estimated.

[0055] Second, considering the technical solution as a whole or from the perspective of the product, the technical effects and advantages of the technical solution to be protected by the present invention are specifically described as follows:

[0056] The present invention proposes a complete method for image recognition in the stability accuracy test, which solves the problem of incomplete image processing in the actual stability accuracy test. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required to be used in the embodiments of the present invention. Obviously, the following described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0058] Figure 1 It is a flowchart of the stability accuracy test method provided by the embodiment of the present invention;

[0059] Figure 2 It is a schematic diagram of the principle of the stability accuracy test method provided by the embodiment of the present invention;

[0060] Figure 3 It is a block diagram of the structure of the stability accuracy test system provided by the embodiment of the present invention;

[0061] In the figure: 1. Original video acquisition module; 2. Target template library construction module; 3. Video processing module; 4. Template selection module; 5. Image processing module; 6. Image feature data calculation module; 7. Stability accuracy numerical calculation module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0062] In order to make the purpose, technical solutions and advantages of the present invention clearer, the following will further describe the present invention in detail with reference to the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0063] In view of the problems existing in the prior art, the present invention provides a stability accuracy test method, system, medium, device and terminal. The following will describe the present invention in detail with reference to the drawings.

[0064] I. Explanation of the embodiments. This part is an explanatory embodiment that expands on the technical solution of the claims to enable those skilled in the art to fully understand how the present invention is specifically implemented.

[0065] As Figure 1 shown, the stable accuracy testing method provided by the embodiments of the present invention includes the following steps:

[0066] S101: Collect the original video data output when the optoelectronic camera on the stable platform tracks the target target.

[0067] S102: Capture a series of single-frame images, and perform segmentation and binarization processing on the images respectively to obtain a target template library; at the same time, intercept an appropriate length of the original video segment, and process the video by removing background noise and local magnification.

[0068] S103: Select the target template with the best effect as the template, and at the same time extract each frame image of the video, and perform segmentation and binarization processing on the images respectively.

[0069] S104: Use the template matching method to match the template image block with the input image, find the most suitable matching part and perform normalization processing, and analyze and calculate to obtain the feature data of the image.

[0070] S105: Substitute the field of view angle, and calculate the stable accuracy value of the entire video segment according to the formula.

[0071] The schematic diagram of the stable accuracy testing method provided by the embodiments of the present invention is as Figure 2 shown.

[0072] As Figure 3 shown, the stable accuracy testing system provided by the embodiments of the present invention includes:

[0073] An original video acquisition module 1 for collecting the original video data output when the optoelectronic camera on the stable platform tracks the target target;

[0074] A target template library construction module 2 for capturing a series of single-frame images and performing segmentation and binarization processing on the images respectively to obtain a target template library;

[0075] A video processing module 3 for intercepting an appropriate length of the original video segment and processing the video by removing background noise and local magnification;

[0076] A template selection module 4 for selecting the target template with the best effect as the template;

[0077] An image processing module 5 for extracting each frame image of the video and performing segmentation and binarization processing on the images respectively;

[0078] An image feature data calculation module 6, which is used to use the template matching method to match the template image block with the input image, find the most suitable matching part and perform normalization processing, and analyze and calculate the feature data of the image;

[0079] A stable precision numerical calculation module 7, which is used to substitute the field of view angle and calculate the stable precision value of the entire video according to the formula.

[0080] The technical solution of the present invention will be further described below in conjunction with specific embodiments.

[0081] The present invention proposes a complete method for image recognition in the stable precision test, which solves the problem of incomplete image processing in the actual stable precision test. The present invention collects the original video of the optoelectronic camera, processes the image to obtain a template, uses the template matching method to recognize the video, obtains the feature data of each frame of the image in the video, and then uses the formula to calculate the stable precision data in the entire video.

[0082] The stable precision test method provided by the embodiment of the present invention specifically includes:

[0083] The optoelectronic camera on the stable platform tracks the relatively moving target target, and the original video output by the collected optoelectronic camera is divided into two paths for subsequent work. One is to capture a single-frame image in the video, and through image segmentation and binarization processing (using the combination of the two functions THRESH_OTSU and THRESH_BINARY), a series of target templates are obtained to form a target template library. The images in the target template library are compared, and the optimal target template is selected as the template for template matching. The other is to select the video segment to be calculated according to the scheme and divide the video segment into single-frame images.

[0084] Using the template matching method (matchTemplate function), by sliding the template image on each frame of the target image and comparing them one by one, using the TM_SQDIFF_NORMED function for image matching, finding the most suitable matching part, after normalization processing, using the minMaxLoc function to find the matching results and their positions of the maximum and minimum values in the matrix, and obtaining the feature data (localized coordinates) recognized by the template in the single-frame image. Substitute the known field of view angle of the optoelectronic camera into the formula for calculation to obtain the stable precision data of the entire segment.

[0085] The displacement deviation of the center of the cross target in two adjacent frames of the video image is denoted as Δx i 、Δy i , calculate 1 stable precision value per second, and count the average value of 1 min.

[0086] The stable precision per second is calculated according to the following formula:

[0087]

[0088]

[0089]

[0090] Δx i —— The number of pixel deviations in the azimuth of the crosshair center in two frames of images;

[0091] Δy i —— The number of pixel deviations in the pitch of the crosshair center in two frames of images;

[0092] θ x —— The measured azimuth field of view angle of the small field of view, °;

[0093] θ y —— The measured pitch and roll field of view angle of the minimum field of view, °;

[0094] X - Resolution in the azimuth direction of the field of view;

[0095] Y - Resolution in the pitch direction of the field of view;

[0096] μ xt —— Azimuth stability accuracy, mrad;

[0097] μ yt —— Pitch stability accuracy, mrad;

[0098] μ t —— Stability accuracy per second, mrad;

[0099] n - Total number of frames.

[0100] Stability accuracy:

[0101]

[0102] In the stability accuracy processing of the present invention, the template matching method (matchTemplate function) is first used to identify the effective features of the image, and the accuracy is accurate to the pixel level and accurate. The accuracy used by others for image recognition is sub - pixel level, and the data at this level is estimated.

[0103] II. Application embodiments. To prove the creativity and technical value of the technical solution of the present invention, this part is an application embodiment of applying the technical solution of the claims to a specific product or related technology.

[0104] The application embodiment of the present invention provides a computer device, the computer device includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor is caused to execute the following steps:

[0105] By collecting the original video image data of an optoelectronic camera and processing the images to obtain templates; after respectively segmenting and binarizing the images, using the template matching method to identify the video, obtaining the feature data of each frame of the image in the video, and then using a formula to calculate the stability accuracy data of the entire video.

[0106] An application embodiment of the present invention provides a computer-readable storage medium storing a computer program, which when executed by a processor causes the processor to perform the following steps:

[0107] By collecting the original video image data of an optoelectronic camera and processing the images to obtain templates; after respectively segmenting and binarizing the images, using the template matching method to identify the video, obtaining the feature data of each frame of the image in the video, and then using a formula to calculate the stability accuracy data of the entire video.

[0108] An application embodiment of the present invention provides an information data processing terminal for implementing the stability accuracy test system described above.

[0109] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those of ordinary skill in the art can understand that the above devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code is provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuits of programmable hardware devices such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips and transistors, or programmable logic devices such as field programmable gate arrays, and can also be implemented by software executed by various types of processors, or can be implemented by a combination of the above hardware circuits and software such as firmware.

[0110] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention by those skilled in the art within the technical scope disclosed by the present invention shall be covered by the protection scope of the present invention.

Claims

1. A stable accuracy test method, characterized in that, The stable accuracy test method includes the following steps: Step 1: Collect the original video data output by the optoelectronic camera on the stable platform when tracking the target target. Step 2: Capture a series of single-frame images from the original video data, and perform segmentation and binarization processing on the images respectively to obtain a target template library; at the same time, intercept an appropriate length of the original video segment, and process the original video segment by removing background noise and local magnification. Step 3: Select the target template with the best effect as the template image, and at the same time extract each frame image of the original video segment, and perform segmentation and binarization processing on the images respectively. Step 4: Use the template matching method to match the template image with the input target image, find the most suitable matching part and perform normalization processing, and analyze and calculate the characteristic data of the target image. Step 5: Substitute the field of view angle of the optoelectronic camera, and calculate the stable accuracy value of the entire original video segment according to the formula. The substituting the field of view angle of the optoelectronic camera in Step 5 and calculating the stable accuracy value of the entire original video segment includes: Δx i represents the number of pixel azimuth deviations of the crosshair center in two frames of images, and Δy i represents the number of pixel pitch deviations of the crosshair center in two frames of images. One stable accuracy value is calculated per second, and the average value of the stable accuracy values for 1 minute is statistically calculated; The stable accuracy value per second is calculated according to the following formula: Among them, θ x represents the measured azimuth field of view angle of the minimum field of view, unit °; θ y represents the measured elevation field of view angle of the minimum field of view, unit °; X represents the resolution in the azimuth direction of the field of view; Y represents the resolution in the elevation direction of the field of view; μ xt represents the azimuth stability accuracy, unit mrad; μ yt represents the elevation stability accuracy, unit mrad; μ t represents the stability accuracy per second, unit mrad; n represents the total number of frames; The stable accuracy is:

2. The stable accuracy test method according to claim 1, characterized in that The stable accuracy test method also includes: The optoelectronic camera on the stable platform tracks the relatively moving target target, and the original video output by the collected optoelectronic camera is divided into the following two paths for subsequent work:

1. Capture single-frame images in the original video, and by performing segmentation and binarization processing on the images, combine the two functions of THRESH_OTSU and THRESH_BINARY to obtain a series of target templates, forming a target template library; compare the images in the target template library, and select the optimal target template as the template image for template matching.

2. Select the video segment to be calculated in the original video according to the scheme, and divide the video segment into single-frame target images.

3. The stable precision testing method according to claim 1, characterized in that The stable accuracy test method also includes: Use the template matching method, utilize the matchTemplate function, slide the template image one by one on each frame of the target image for comparison, use the TM_SQDIFF_NORMED function for image matching, find the most suitable matching part, and perform normalization processing. Use the minMaxLoc function to find the matching results and positions of the maximum and minimum values after normalization processing, and obtain the characteristic data or positioning coordinates recognized by the template image in the single-frame target image; substitute the known field of view angle of the optoelectronic camera into the formula for calculation to obtain the stable accuracy data of an entire original video segment.

4. The stable precision test method according to claim 1, characterized in that The performing segmentation and binarization processing on the images respectively in Step 2 includes: Receive a series of single-frame images in the captured original video data, obtain the class label and binarized segmentation label of the single-frame image, and perform feature extraction on the single-frame image to obtain a training data set. Obtain a target heat map through a preset deep learning convolutional neural network, and according to the target heat map and the binarized segmentation label, obtain an image segmentation loss value. Adjust the parameters of the preset deep learning convolutional neural network according to the training data set and the image segmentation loss value, and then complete the training of the model. Using the trained deep learning convolutional neural network model, perform segmentation and binarization processing on a series of single-frame images in the captured original video data respectively, and output the processing results.

5. A stable accuracy test system applying the stable accuracy test method according to any one of claims 1 to 4, characterized in that, The stable accuracy test system includes: An original video acquisition module, configured to acquire the original video data output by an optoelectronic camera on a stable platform when tracking a target; A target template library construction module, configured to capture a series of single-frame images in the original video data, and perform segmentation and binarization processing on the images respectively to obtain a target template library; A video processing module, configured to intercept an original video segment of an appropriate length, and process the original video segment by removing background noise and local magnification; A template selection module, configured to select the target template with the best effect as the template image; An image processing module, configured to extract each frame image of the original video segment, and perform segmentation and binarization processing on the images respectively; An image feature data calculation module, configured to use the template matching method, match the template image with the input target image, find the most suitable matching part and perform normalization processing, and analyze and calculate the feature data of the target image; A stable accuracy numerical calculation module, configured to substitute the field of view angle of the optoelectronic camera, and calculate the stable accuracy numerical value of the entire original video segment according to the formula.

6. A computer device, characterized in that, The computer device includes a memory and a processor. When the computer program stored in the memory is executed by the processor, the processor is caused to execute the steps of the stable accuracy test method as described in claim 1.

7. A computer-readable storage medium storing a computer program, which when executed by a processor, causes the processor to execute the steps of the stable accuracy test method as described in claim 1.

8. An information data processing terminal, characterized in that The information data processing terminal is used to implement the stable accuracy test system as described in claim 5.

Images