Substation monitoring camera image supplementary screening method based on spatial size

Abstract

The invention relates to the technical field of transformer substation remote vision systems, and particularly discloses a transformer substation monitoring camera image supplementing and screening method based on a spatial size, which comprises the following steps: acquiring an image from a camera, determining the size level of an observation object in the image, selecting a closest observation reference from an observation reference library according to the size level of the observation object, and screening the image. Scoring parameters are obtained, the images are scored according to the observation benchmark, and the images are screened according to scores. Compared with the prior art, the image screening method provided by the invention solves the problems that an image screening means mainly depends on manpower and is low in efficiency, and the image screening method provided by the invention avoids image analysis from a pixel level and judgment from a macroscopic distance in space, and is simpler and more direct.

Description

Technical Field

[0001] The invention belongs to the technical field of substation remote viewing systems, and in particular relates to a substation monitoring camera image supplement screening method based on space size. Background Art

[0002] Through the substation monitoring camera, operation and maintenance personnel can remotely view the equipment. In the substation, the same equipment can often be observed by multiple cameras at the same time. Affected by factors such as shooting angle, distance, and camera focal length, the image quality captured by different cameras varies. It may even be impossible to identify the details of the equipment in the captured image, and only the general outline can be seen. Therefore, it is necessary to screen out suitable images for display.

[0003] At present, the image screening method of the substation remote viewing system mainly relies on manual work, and the operation and maintenance personnel need to visually distinguish each image transmitted back. When the number of cameras is small, the workload is relatively small and can be covered by manpower; if there are many cameras, the images are transmitted to the operation and maintenance center without screening, which seriously occupies the tight signal transmission channel resources, and the operation and maintenance personnel are required to screen them, which is time-consuming and labor-intensive, and is not conducive to improving work efficiency.

[0004] At present, commonly used objective image quality evaluation methods can be generally divided into three categories, namely full reference image quality evaluation, semi-reference image quality evaluation and no reference image quality evaluation. These methods can identify the clarity and distortion level of the image. However, in the special environment of the substation, the staff pays more attention to the main body of the equipment. If an image is evaluated only from one level such as clarity, the image may be clear and distortion-free, but the object to be observed in the image is too small to see the details of the equipment. Summary of the invention

[0005] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a method for supplementing and screening substation monitoring camera images based on spatial dimensions.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] The present invention provides a method for supplementing and screening images of substation monitoring cameras based on spatial dimensions, comprising the following steps: acquiring an image from a camera, determining the size level of an observed object in the image, selecting the closest observation benchmark from an observation benchmark library according to the size level of the observed object, acquiring a scoring parameter, scoring the image according to the observation benchmark, and screening the image according to the score.

[0008] Furthermore, the method for establishing the observation reference library comprises the following steps:

[0009] Step S1: Establish a standard camera shooting scene, adjust the camera focal length to the minimum, and define the minimum focal length as f min , wherein the photographed object is a regular cube, a mark circle is engraved on the surface of the regular cube, and a diameter is inside the mark circle, and the length of the diameter is the size d0 of the observation reference;

[0010] Step S2: Move the object until its diameter cannot be distinguished, and record the shooting distance L max ;

[0011] Step S3: Calculate the pixel space length of the image diameter based on the Hough line change, and calculate the pixel point size D0 in combination with the size d0 of the observation reference;

[0012] Step S4: Change the diameter d0 of the circle and execute step S1 until the maximum size of all observation devices is covered, and establish an observation benchmark library based on the data obtained from multiple operations.

[0013] Furthermore, the shooting distance L in step S2 is max is the straight-line distance between the camera and the object.

[0014] Furthermore, the calculation formula of the pixel size D0 is:

[0015]

[0016] where d px is the pixel space length, and α is the angle between the object and the imaging plane of the camera.

[0017] Furthermore, the data of step S4 includes the shooting distance L max , the size of the observation reference d0, the pixel size D0 and f min .

[0018] Furthermore, the scoring parameters include the distance between the photographed object and the corresponding camera, and the shooting focal length of the corresponding camera.

[0019] Furthermore, the image is scored according to the following scoring rules:

[0020] The distance between the object and the corresponding camera and the selected observation reference l max The value is compared, if it is greater than L max value, then the distance between the photographed object and the corresponding camera and the selected observation benchmark L fmax The value is compared, if it is greater than l fmax value, set the image score to -1, if it is less than or equal to L fmax value, set the image score to 0, if it is less than or equal to L maxvalue, then score according to the scoring formula.

[0021] Furthermore, the scoring formula is:

[0022]

[0023] Among them, S is the image score, f is the focal length of the corresponding camera, d0 is the size of the observation reference, and d px is the pixel space length, α is the angle between the object and the imaging plane of the camera, l is the distance between the object and the corresponding camera, L fmax is the maximum focal length.

[0024] Furthermore, the L fmax =D0f max .

[0025] Furthermore, the specific steps of image screening are as follows: directly discard images with a score of -1, sort images with a score greater than 0 according to the score, if the number of images after sorting meets the task requirements, discard images with a score less than or equal to 0, if not, add images with a score of 0 to the sort, if the number of images after sorting meets the task requirements, discard images with a score less than 0, if not, adjust the focal length of the camera corresponding to the image with a score greater than -1, and continue to score and screen according to the scoring formula until the observation task is completed.

[0026] Compared with the prior art, the present invention.

[0027] (1) The present invention solves the problem that image screening methods mainly rely on manual labor and are inefficient.

[0028] (2) The establishment of an observation benchmark library can classify and organize observation tasks for different objects, greatly reducing the work of pixel size calculation;

[0029] (3) Directly evaluating image quality based on spatial size avoids analyzing image quality at the level of pixel space, greatly reducing the difficulty of image quality evaluation.

[0030] (4) The image screening method proposed by the present invention makes judgments based on the macroscopic distance in space, which is simpler and more direct. The image screening method proposed by the present invention avoids analyzing the image at the pixel level and makes judgments based on the macroscopic distance in space, which is simpler and more direct. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a schematic diagram of the shooting distance at the minimum focal length;

[0032] Figure 2 This is a shooting diagram for adjusting the focal length;

[0033] Figure 3 It is the image screening flow chart. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0035] Pixel size refers to the actual physical space that a single pixel occupies on the screen. It determines the number of pixels that can be accommodated per unit length on the display. Smaller pixel size means more pixels per unit length on the display, thereby improving the resolution and clarity of the image.

[0036] The present invention establishes the following evaluation rules: taking the minimum focal length as the basic premise for testing image quality. The reason is that if the image meets the requirements at the minimum focal length, then the image after the focal length is magnified will meet the requirements. If a certain observed object needs to be clearly observed after the focal length is magnified, the adjustable range of the camera's focal length will be reduced accordingly, so its priority is lowered. If the focal length is adjusted to the maximum value and does not meet the observation requirements, the camera cannot be used to observe the object and must be discarded.

[0037] The present invention uses a regular cube as the shooting object, fixes the camera position, determines the camera resolution, and adjusts the camera to the minimum focal length f min , respectively test the clarity of the cube at different distances, and obtain the image quality evaluation criteria based on pixel size. The specific method is: a circle is drawn on the surface of the cube, and a diameter is drawn inside the circle. The length of the diameter is the size of the observation reference d0. The cube is translated along the shooting direction, and the direction of the diameter at different distances is observed. The distance until the diameter of the photographed object cannot be distinguished is the maximum shooting distance L. max , where whether the image diameter can be distinguished is determined by using an edge detection algorithm to extract the diameter edge in the image and calculating the edge strength value. Shooting distance L max is the straight-line distance between the camera and the object being photographed. At the same time, the pixel space length d of the diameter is solved based on the Hough straight line transform px Therefore, in order to meet the observation requirements, the standard calculated value of the pixel size that needs to be achieved is as follows:

[0038]

[0039] The shooting distance corresponding to the pixel size D0 is L maxBased on the principle of similar triangles, if the size of the observed object is different, under the premise of the minimum shooting focal length, in order to maintain the observation standard of pixel size D0 unchanged, such as Figure 1 Figure 2 The calculation result is shown in formula 2.

[0040]

[0041] Where d is the size of the observed object, d ′ px is the pixel space length, L is the shooting distance, and Formula 2 is rearranged as shown in Formula 3 below.

[0042] L≤L max (3)

[0043] From Formula 3, we can see that in order to ensure that the pixel size D0 standard can always be met, the shooting distance L only needs to be no more than L max That is, it has nothing to do with the size parameters of the observed object. Therefore, under the premise that the shooting distance of different observed object sizes d satisfies Formula 3, the image meets the requirements.

[0044] When the shooting distance exceeds the value calculated by formula 3, the focal length needs to be adjusted to make the image meet the requirements. The calculation method is as follows.

[0045]

[0046] It is further calculated that the shooting distance L at different focal lengths f should satisfy Formula 5.

[0047] L≤D0f(5)

[0048] When f is the maximum focal length f max When the maximum shooting distance is L fmax =D0f max .

[0049] From the above formula 5, we can know that if we observe objects of different sizes at different distances, and the shooting distance L>L max When L does not exceed d0f, the image meets the requirements, which is also independent of the observation size. It is worth noting that when f = f min When L≤L max , so Formula 5 can be considered as the core criterion for determining image quality.

[0050] From the above calculation process, we can know that after determining the standard value D0 of the pixel size of the evaluation image, combined with the imaging principle of the camera, the pixel size is used as an intermediate link to realize the use of shooting distance and focal length as key parameters for evaluating image quality. Compared with the process of obtaining pixel size, the difficulty and complexity of the acquisition method of parameters such as shooting distance and focal length are significantly reduced. When the shooting distance is determined at the camera position, this value is a fixed value; the focal length is the camera's own parameter and can be directly obtained.

[0051] Therefore, the present invention sets the maximum shooting distance L at the minimum focal length max The ratio between the image and the shooting distance L is used as a parameter for scoring the image. Its physical meaning is that when the distance does not exceed L max Under the premise of L, the closer the shooting distance, the higher the image score and the better the quality. max If the image score is 0, the image is discarded. If the camera needs to be used, the focal length needs to be increased, and the adjustable focal length range of the camera will be reduced simultaneously, which also reflects that the "observation potential" of the camera is relatively small. Therefore, the present invention downgrades the camera image and scores it. fmax , the image score is -1, and the camera cannot observe the object clearly through adjustment.

[0052] It is worth noting that if there is a certain angle α between the reference observation object d0 and the imaging plane of the camera, if the object needs to be observed clearly, it is equivalent to observing a new object facing the camera, and the size of the object is d0cosα. Therefore, the calculation formula of the pixel size is shown in Formula 6.

[0053]

[0054] The larger the angle α is, the smaller the pixel size is. When α = 90, dcosα is 0, D0 = 0, and its physical meaning is that the actual object length corresponding to a unit pixel is 0, and the object cannot be observed. Therefore, based on the evaluation rules described above, the coefficient of cosα needs to be increased when scoring the image. The calculation method of the image score S is as follows.

[0055]

[0056] In fact, in the substation environment, the observation objects are diverse. Sometimes it is to observe whether there is a large area of ​​dirt on the insulator, sometimes it is to observe the reading or the indicated area on the meter, and sometimes it is to observe whether the moving and static contacts of the disconnector are burned, etc. For different observation objects, the required pixel size standard D0 value is different. For example, for dirt, it is not necessary to look at the situation inside the dirt, but only to observe the size of the area covered by the dirt on the insulator. A small amount of dirt has little effect on the insulation performance. When the dirt area accumulates to a certain amount, it needs to be observed. Therefore, its observation size is relatively large, and the spacing of an insulator shed can be used as the observation benchmark; for meter readings, staff often care about the area it points to or even the precise reading, so its observation size benchmark is the size of the scale number on the meter. In summary, different types of pixel size benchmarks need to be set for different situations for reference, and the same standard can be used for observation objects of similar sizes. Generally speaking, an observation benchmark library can be established to cover most observation services in substation operation and maintenance. For example, observation standards can be divided into 1 mm, 5 mm, 1 cm, 5 cm, 10 cm and other types.

[0057] The image screening method proposed by the present invention avoids analyzing the image at the pixel level and makes judgments based on the macroscopic distance in space, which is simpler and more direct. The image screening method proposed by the present invention avoids analyzing the image at the pixel level and makes judgments based on the macroscopic distance in space, which is simpler and more direct.

[0058] A complementary screening method for substation surveillance camera images based on spatial dimensions, such as Figure 3 As shown, including:

[0059] Step 1: Establishment of observation benchmark database;

[0060] Step 2, select the observation benchmark;

[0061] Step 3, obtain scoring parameters;

[0062] Step 4: Image scoring and screening.

[0063] In addition, the substation monitoring camera image supplementary screening method based on spatial size provided by the present invention also has the following technical features:

[0064] Furthermore, step 1 specifically includes:

[0065] Step 1.1, build an experimental platform and set up a standard camera shooting scene, where the shooting object is a regular cube with a circle engraved on the surface of the cube. There is a diameter inside the circle. The length of the diameter is the size of the observation reference d0. The minimum camera focal length adjustment value is f min ;

[0066] Step 1.2: Move the subject until the diameter of the subject cannot be distinguished, and record the shooting distance L. max , shooting distance L max is the straight-line distance between the camera and the object;

[0067] Step 1.3, solving the pixel size, solving the pixel space length of the inner diameter of the image based on the Hough line change, and combining the size d0 of the observation reference to calculate the pixel size D0;

[0068] Step 1.4, establish the observation reference library, change the circle diameter d0 value, repeat steps 1.1-1.3, calculate the new pixel size reference, and record the corresponding shooting distance L max value and establish an observation benchmark library.

[0069] Furthermore, step 2 specifically includes: determining the size level of the observed object, and selecting the closest observation benchmark from the observation benchmark library established in step 1.

[0070] Furthermore, step 3 specifically includes:

[0071] Step 3.1, obtaining the shooting distance. The surveillance cameras and equipment in the substation are fixed objects, and their positions do not change. The distance between the observed object and the corresponding camera is measured;

[0072] Step 3.2, record the camera focal length.

[0073] Furthermore, step 4 specifically includes:

[0074] Step 4.1, determine the image priority, according to the shooting distance in step 3.1, if it exceeds L max value, the priority is reduced by one level and the image score is set to 0. fmas value, the priority is the lowest and the image score is set to -1;

[0075] Step 4.2, image sorting, sort the images in step 4.1 with a shooting distance no greater than L max The images with the same value are scored according to Formula 7, and the images are sorted in descending order according to the scores.

[0076] Step 4.3, image screening, discard the image with a score of -1 directly, show the sorted images in step 4.2 to the operation and maintenance personnel, and if the task requirements are met, discard the image with a score of 0. Otherwise, adjust the focal length of the camera corresponding to the image with a score greater than -1, and continue to score and screen according to formula 7 until the observation task is completed.

[0077] According to a method for supplementing and screening substation surveillance camera images based on spatial dimensions provided by the present invention, the core purpose is to intuitively evaluate the quality of substation camera images based on spatial dimensions, score the images transmitted by each camera, eliminate images with unqualified quality, and prioritize qualified images according to their scores and display them to staff.

[0078] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for supplementing and screening substation surveillance camera images based on spatial dimensions, characterized in that: The following steps are involved: Acquire an image from the camera, determine the size level of the observed object in the image, select the closest observation benchmark from the observation benchmark library according to the size level of the observed object, obtain scoring parameters, score the image according to the observation benchmark, and filter the image according to the score.

2. The method for supplementing and screening substation surveillance camera images based on spatial dimensions according to claim 1 is characterized in that: The method for establishing the observation reference library comprises the following steps: Step S1: Establish a standard camera shooting scene, adjust the camera focal length to the minimum, and define the minimum focal length as f min , wherein the photographed object is a regular cube, a mark circle is engraved on the surface of the regular cube, and a diameter is inside the mark circle, and the length of the diameter is the size d0 of the observation reference; Step S2: Move the object until its diameter cannot be distinguished, and record the shooting distance L max ; Step S3: Calculate the pixel space length of the image diameter based on the Hough line change, and calculate the pixel point size D0 in combination with the size d0 of the observation reference; Step S4: Change the diameter d0 of the circle and execute step S1 until the maximum size of all observation devices is covered, and establish an observation benchmark library based on the data obtained from multiple operations.

3. The method for supplementing and screening substation surveillance camera images based on spatial dimensions according to claim 2 is characterized in that: The step S2 takes a picture at a distance L max is the straight-line distance between the camera and the object.

4. The method for supplementing and screening substation surveillance camera images based on spatial dimensions according to claim 2 is characterized in that: The calculation formula of the pixel size D0 is: where d px is the pixel space length, and α is the angle between the object and the camera imaging plane.

5. The method for supplementing and screening substation surveillance camera images based on spatial dimensions according to claim 2 is characterized in that: The data of step S4 includes the shooting distance L max , the size of the observation reference d0, the pixel size D0 and f min .

6. The method for supplementing and screening substation monitoring camera images based on spatial dimensions according to claim 1 is characterized in that: The scoring parameters include the distance between the photographed object and the corresponding camera, and the shooting focal length of the corresponding camera.

7. The method for supplementing and screening substation surveillance camera images based on spatial dimensions according to claim 1 is characterized in that: The scoring rules for the image are as follows: The distance between the object and the corresponding camera and the selected observation reference L max The value is compared, if it is greater than L max value, then the distance between the photographed object and the corresponding camera and the selected observation benchmark L fmax The value is compared, if it is greater than L fmax value, set the image score to -1, if it is less than or equal to L fmax value, set the image score to 0, if it is less than or equal to L max value, then score according to the scoring formula.

8. The method for supplementing and screening substation surveillance camera images based on spatial dimensions according to claim 7 is characterized in that: The scoring formula is: Among them, S is the image score, f is the focal length of the corresponding camera, d0 is the size of the observation reference, and d px is the pixel space length, α is the angle between the object and the imaging plane of the camera, l is the distance between the object and the corresponding camera, L fmax is the maximum focal length.

9. The method for supplementing and screening substation monitoring camera images based on spatial dimensions according to claim 8, characterized in that: The L fmax =D0f max .

10. The method for supplementing and screening substation surveillance camera images based on spatial dimensions according to claim 1, characterized in that: The specific steps of image screening are as follows: directly discard images with a score of -1, sort images with a score greater than 0 according to the score, if the number of images after sorting meets the task requirements, discard images with a score less than or equal to 0, if not, add images with a score of 0 to the sort, if the number of images after sorting meets the task requirements, discard images with a score less than 0, if not, adjust the focal length of the camera corresponding to the image with a score greater than -1, and continue to score and screen according to the scoring formula until the observation task is completed.

Images