Method for judging the balance state of the pointer of a tray balance for physical, chemical and biological experiments

Abstract

The present invention discloses a method for judging the balance state of a pan balance pointer for physical, chemical and biological experiments. The method reduces the error caused by abnormal situations by filtering abnormal differential images and screens the final pointer contour using prior conditions, thereby solving the influence of illumination changes and low image contrast on pointer extraction and improving the recognition speed and accuracy of the pointer.

Description

Technical Field

[0001] The invention relates to the technical field of image recognition, and in particular to a method for judging the balance state of a tray balance pointer for physical, chemical and biological experiments. Background Art

[0002] In the teaching of physics, chemistry and biology experiments in middle schools, the use of a pan balance is one of the common experimental operations. Among them, moving the cursor on the ruler until the pointer points to the center of the dial again is an important pre-operation before using the pan balance, which is used to test students' mastery of balance adjustment.

[0003] The traditional scoring of balance leveling experiments mainly relies on teachers' manual observation to judge whether students have leveled the two ends of the balance before weighing, which has problems such as strong subjectivity and low efficiency. With the development of artificial intelligence and computer vision technology, automatic scoring of physics, chemistry and biology experimental operation tests has become possible.

[0004] Chinese invention CN115457556A discloses a method for reading a circular pointer instrument in a nuclear power plant, which uses an image acquisition device such as a camera to obtain an image of a pointer instrument; uses models such as deep learning to detect, identify, and extract the instrument dial, key parameters, and pointer; calculates the relative position of the pointer in the dial to eliminate the reading deviation caused by angle deflection; calculates the current instrument reading based on the relative angle between the pointer and the scale reference line, solves the reading deviation problem caused by the angle tilt of the instrument, the interference problem between the dial pattern and the pointer style, and improves the accuracy of the instrument reading and the recognition efficiency.

[0005] However, the above method determines the starting point and end point of the pointer based on the Euclidean distance formula, and calculates the relative angular position of the pointer on the dial. The linear regression model based on the least squares method fits the pointer straight line to calculate the relative angular position of the pointer on the dial, which will cause a large error in the result due to the slight deviation of the original image. Summary of the invention

[0006] In view of the above problems, the present invention proposes a method for judging the balance state of a pan balance pointer for physical, chemical and biological experiments, aiming to quickly and accurately extract the pan balance pointer and detect its balance state.

[0007] In order to solve the above technical problems, the first aspect of the present invention provides a method for judging the balance state of a tray balance pointer for physical, chemical and biological experiments, comprising the following steps:

[0008] Intercepting an original image of an experimenter adjusting the balance of a tray from a video stream, and intercepting a target image including the tray from the original image;

[0009] Preprocessing the target image to obtain a rectangular positioning frame including a dial and a pointer, and overlaying the rectangular positioning frame onto the original image as a template image;

[0010] Using the rectangular positioning frame to intercept a corresponding local image from the video stream, performing a differential calculation between the local image and the template image to obtain a differential image, counting the motion area of ​​each frame according to the differential image, and filtering abnormal differential images according to preset rules;

[0011] Binarizing the filtered differential image, and then removing the interfering contours in the binary image;

[0012] According to the prior conditions, a pointer contour meeting the preset conditions is selected from the binary image as a final pointer contour;

[0013] The pointer endpoint coordinates of the final pointer outline in each frame of the binary image are recorded, and it is determined whether the change value of the pointer endpoint coordinates within a preset time window is stable within a preset range. If so, a balance determination is performed.

[0014] In some embodiments, the preprocessing includes: performing white threshold segmentation on the target image to obtain a white pallet outline and a white dial outline, relocating the original color dial outline in the target image according to the relative position relationship between the white pallet outline and the white dial outline, using the minimum outer rectangle of the original color dial outline and expanding the rectangle according to the shape ratio of the dial to obtain the rectangular positioning frame including the dial and the pointer, and overlaying the rectangular positioning frame on the original image as the template image.

[0015] In some implementations, the preset rule includes: counting the proportion of the motion area, and if the proportion is greater than a preset proportion, repeating the preprocessing operation, and updating the rectangular positioning frame and the template image.

[0016] In some implementations, before the filtered differential image is binarized, a pixel deviation threshold is preset, and the differential image is binarized using the pixel deviation threshold.

[0017] In some embodiments, the step of screening the pointer contour that meets the preset conditions from the binary image as the final pointer contour according to the prior conditions includes: setting a confidence threshold, setting the prior conditions according to the pointer characteristics, and calculating the confidence of each pointer contour in the binary image according to the prior conditions; if the confidences corresponding to all the pointer contours are less than the confidence threshold, then attenuating the pixel deviation threshold with a preset step size, and returning to the step of removing the interfering contours in the binary image until multiple candidate contours are obtained, and taking the pointer contour with the highest confidence among the candidate contours as the final pointer contour.

[0018] In some embodiments, the prior conditions include pointer shape, length and / or orientation.

[0019] In some embodiments, the balance determination includes: presetting a dial area as a balance interval of the balance, and if the coordinates of the pointer endpoint fall into the balance interval, it is determined that the balance is balanced at the current moment.

[0020] The second aspect of the present invention proposes a system for judging the balance state of a tray balance pointer for physical, chemical and biological experiments, comprising:

[0021] The original image acquisition module is used to intercept the original image of the experimenter adjusting the balance of the tray from the video stream, and intercept the target image containing the tray from the original image;

[0022] A template image generation module is used to pre-process the target image, obtain a rectangular positioning frame including a dial and a pointer, and overlay the rectangular positioning frame onto the original image as a template image;

[0023] An abnormal image filtering module is used to intercept a corresponding local image from the video stream using the rectangular positioning frame, perform a differential calculation between the local image and the template image to obtain a differential image, count the motion area of ​​each frame according to the differential image, and filter the abnormal differential image according to a preset rule;

[0024] A noise filtering module, used for binarizing the filtered differential image and then removing interference contours in the binary image;

[0025] A pointer contour recognition module, used for selecting a pointer contour meeting a preset condition from the binary image according to a priori conditions as a final pointer contour;

[0026] The pointer balance recognition module is used to record the pointer endpoint coordinates of the final pointer outline in each frame of the binary image, and determine whether the change value of the pointer endpoint coordinates in a preset time window is stable within a preset range. If so, a balance determination is performed.

[0027] The third aspect of the present invention proposes a device for judging the balance state of a tray balance pointer for physical, chemical and biological experiments, the device comprising a memory, a processor and a communication module, wherein:

[0028] The memory is used to store executable program code;

[0029] The processor is coupled to the memory;

[0030] The processor calls the executable program code stored in the memory to execute the above-mentioned method for judging the balance state of the pointer of the tray balance for physical, chemical and biological experiments.

[0031] A fourth aspect of the present invention provides a computer storable medium storing computer instructions, which, when called, are used to execute the above-mentioned method for determining the balance state of a pan balance pointer for physical, chemical and biological experiments.

[0032] The beneficial effects of the present invention are as follows: by filtering abnormal differential images, the errors caused by abnormal situations are reduced, and by using prior conditions to screen the final pointer contour, the influence of illumination changes and low image contrast on pointer extraction is solved, thereby improving the recognition speed and accuracy of the pointer. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A schematic flow chart of a method for determining the balance state of a pointer of a tray balance for physical, chemical and biological experiments disclosed in an embodiment of the present invention;

[0034] Figure 2a The original image of the experimenter adjusting the balance of the pan is captured from the video stream;

[0035] Figure 2b To capture a target image containing a tray from an original image;

[0036] Figure 3a It is a schematic diagram of the white tray outline and the white dial outline;

[0037] Figure 3b Schematic diagram extracted from the template graph;

[0038] Figure 4a Schematic diagram of the final pointer contour extraction;

[0039] Figure 4b A schematic diagram for determining whether the pointer endpoint falls within the equilibrium interval. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solution and advantages of the present invention clearer and more specific, the content of the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. It is understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for the convenience of description, only the parts related to the present invention are shown in the accompanying drawings, rather than all the contents.

[0041] Embodiment 1

[0042] This embodiment proposes a method for judging the balance state of a tray balance pointer for physical, chemical and biological experiments. It adopts an efficient abnormal target screening strategy to reduce the amount of calculation, so that the system can quickly process images and provide real-time and accurate feedback on the balance state of the balance, meeting the needs of large-scale experimental scoring.

[0043] like Figure 1 As shown, the following steps are included:

[0044] S1, captures the original image of the experimenter adjusting the balance of the tray from the video stream, and captures the target image containing the tray from the original image.

[0045] Specifically, the detection module constructed by the neural network model recognizes the pan balance contained in the original image. The original image of the experimenter adjusting the balance of the pan is as follows: Figure 2a As shown, it should be noted that only when the experimenter's hands are close to the pan balance at a certain distance at the same time, it will be judged that the experimenter is trying to adjust the pan balance and the position information of the pan balance will be output. At this time, a rectangular frame ( Figure 2a The red box in the image is used to define the range of the pallet balance and mark the area as a pallet. Then, the image of the pallet balance is intercepted from the original image as the target image. The target image is shown in FIG. Figure 2b As shown. Reduce the size of the image to be processed and prepare for the rapid subsequent image processing.

[0046] The above detection module can be constructed in advance by a technician and given a suitable number of samples, weights and bias parameters for training to obtain a prediction performance that satisfies the corresponding model, with the goal of correctly identifying the experimenter's hands approaching the pan balance at the same time.

[0047] S2, preprocessing the target image, obtaining a rectangular positioning frame including the dial and the pointer, and overlaying the rectangular positioning frame on the original image as a template image.

[0048] In one example, the preprocessing includes: performing white threshold segmentation on the target image to obtain a white tray outline and a white dial outline, and except for the tray and dial area, the rest of the area is a black background. Figure 3aAs shown, the preprocessing method using black and white threshold segmentation consumes less system resources, can effectively simplify image data, highlight key information, and facilitate subsequent image analysis and processing. At the same time, black and white threshold segmentation can also be used as a means of denoising. By selecting an appropriate threshold, the noise in the image (such as small spots, noise, etc.) can be filtered out to obtain a clearer image. Then, according to the relative position relationship between the white tray outline and the white dial outline, the original color dial outline is relocated in the target image, and the minimum outer rectangle of the original color dial outline is used and expanded in proportion to the shape of the dial to obtain a rectangular positioning frame containing the dial and the pointer. The rectangular positioning frame is used as the region of interest (ROI), and the rectangular positioning frame is overlaid on the original image as a template image. The positioning schematic diagram of the template image is shown in the figure. Figure 3b As shown, the green box in the template diagram is the rectangular positioning box containing the dial and the pointer.

[0049] S3, using a rectangular positioning frame to capture the corresponding local image from the video stream, performing differential calculation between the local image and the template image to obtain a differential image, counting the motion area of ​​each frame according to the differential image, and filtering abnormal differential images according to preset rules.

[0050] The application scenario of this embodiment is physical, chemical and biological experiments. In this scenario, the camera is not allowed to move. Therefore, the rectangular positioning frame obtained in S2 can be extended to the video stream. There is no need to re-identify the position of the dial and the pointer for each frame of the image, which can further reduce the consumption of system resources and speed up the image processing.

[0051] In one example, the preset rules include: counting the proportion of the motion area, if the proportion is greater than the preset proportion, repeating the preprocessing operation in S2, updating the rectangular positioning frame and the template image, which can avoid the template image initialization abnormality caused by the displacement, occlusion or image decoding abnormality of the pallet balance, filter out the situation of large movement, and only retain the situation of small movement when the pointer rotates. Even in the case of unstable lighting conditions and low image contrast, the pointer of the pallet balance can be accurately identified, thereby improving the accuracy of pointer recognition.

[0052] S4, the filtered difference image is binarized, and then the interfering contours in the binary image are removed.

[0053] Before the filtered differential image is binarized, it also includes presetting a pixel deviation threshold, using the pixel deviation threshold to binarize the differential image, corroding and filtering the contour, and removing irrelevant small contours to prepare for S5.

[0054] In S4, by setting an appropriate pixel deviation threshold, the target can be accurately separated from the background and the influence of noise and interference can be reduced. This helps to improve the accuracy and completeness of target detection and prepare for the subsequent final pointer contour recognition task.

[0055] S5, selecting a pointer contour that meets a preset condition from the binary image according to a priori conditions as a final pointer contour.

[0056] In one example, a confidence threshold is set, a priori conditions are set according to pointer features, and the confidence of each pointer contour in the binary image is calculated according to the priori conditions. If the confidences corresponding to all pointer contours are less than the confidence threshold, the pixel deviation threshold is attenuated with a preset step size, and the process returns to S4 to remove interfering contours in the binary image until multiple candidate contours are obtained, and the pointer contour with the highest confidence among the candidate contours is taken as the final pointer contour. In this step, the pixel deviation threshold is adaptively and iteratively attenuated, and automatic adjustment can be made according to different lighting and contrast conditions, thereby improving the pointer recognition accuracy in experimental environments with unstable lighting conditions and many image interference factors, thereby ensuring that a reliable final pointer contour can be obtained even in a changing experimental environment.

[0057] Since this solution needs to extract multiple original images from the video stream, a corresponding number of binary images and candidate contours will also be generated, from which the pointer contour with the highest confidence is extracted as the final pointer contour.

[0058] The above-mentioned prior conditions include pointer shape, position and / or orientation. The calculation of confidence is obtained by quantifying the degree of conformity between the candidate contour and the prior conditions, and summing them up according to certain weights. The specific effect is that the contour located in the middle of the template image, with a slender contour and a contour orientation angle close to vertical can be preferentially extracted as the pointer contour. The larger the height / width ratio of the circumscribed rotation box of the candidate contour, the slenderer the pointer shape, and the higher the degree of conformity of the candidate contour; the closer the center point of the circumscribed rotation box of the candidate contour is to the x-coordinate of the image center, the higher the degree of conformity; the closer the orientation angle of the candidate contour endpoint is to 90°, the higher the degree of conformity, etc. The final pointer contour extracted is as follows: Figure 4a shown.

[0059] S6, recording the pointer endpoint coordinates of the final pointer outline in each frame of the binary image, and determining whether the change value of the pointer endpoint coordinates within the preset time window is stable within a preset range. If so, performing a balance determination.

[0060] Balance determination includes: presetting a dial area as the balance interval of the balance, and the balance interval can refer to Figure 4b If the two parallel red dotted lines are set, the pointer endpoint coordinates fall into the balance interval (such as Figure 4bAs shown, the final pointer outline is represented by a green box, the pointer endpoint is the red point marked within the green box, and this pointer endpoint is within the balance interval), then it is judged that the balance is balanced at the current moment.

[0061] In this embodiment, reliable pointer balance recognition results are obtained mainly by relying on the filtering operations and screening operations of S3 and S5.

[0062] Embodiment 2

[0063] This embodiment proposes a system for judging the balance state of a tray balance pointer for physical, chemical and biological experiments, including:

[0064] The original image acquisition module is used to intercept the original image of the experimenter adjusting the balance of the tray from the video stream, and intercept the target image containing the tray from the original image;

[0065] The template image generation module is used to pre-process the target image, obtain a rectangular positioning frame including the dial and the pointer, and overlay the rectangular positioning frame onto the original image as a template image;

[0066] The abnormal image filtering module is used to intercept the corresponding local image from the video stream using a rectangular positioning frame, perform differential calculation between the local image and the template image to obtain a differential image, count the motion area of ​​each frame according to the differential image, and filter the abnormal differential image according to preset rules;

[0067] A noise filtering module is used to binarize the filtered difference image and then remove the interfering contours in the binary image;

[0068] The pointer contour recognition module is used to select the pointer contour that meets the preset conditions from the binary image according to the prior conditions as the final pointer contour;

[0069] The pointer balance recognition module is used to record the pointer endpoint coordinates of the final pointer outline in each frame of the binary image, and determine whether the change value of the pointer endpoint coordinates in the preset time window is stable within the preset range. If so, a balance judgment is made.

[0070] The functions implemented by each module of this embodiment can refer to the method described in the first embodiment, corresponding to steps S1-S6, and will not be described in detail later.

[0071] Embodiment 3

[0072] The embodiment of the present invention discloses a device for judging the balance state of a tray balance pointer for physical, chemical and biological experiments. The device may include:

[0073] A memory storing executable program code;

[0074] a processor coupled to the memory;

[0075] The processor calls the executable program code stored in the memory to execute the steps of the method for determining the balanced state of a pointer of a tray balance for physical, chemical and biological experiments described in the first embodiment.

[0076] Embodiment 4

[0077] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program can be operated to enable a computer to execute the steps in the method for determining the balance state of a pan balance pointer for physical, chemical and biological experiments described in Example 1.

[0078] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0079] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0080] Finally, it should be noted that the method for judging the balance state of a pan balance pointer for physical, chemical and biological experiments disclosed in the embodiment of the present invention only discloses a preferred embodiment of the present invention, which is only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments.

Claims

1. A method for judging the balance state of a pan balance pointer for physical, chemical and biological experiments, characterized in that: The following steps are involved: Intercepting an original image of an experimenter adjusting the balance of a tray from a video stream, and intercepting a target image including the tray from the original image; Preprocessing the target image to obtain a rectangular positioning frame including a dial and a pointer, and overlaying the rectangular positioning frame onto the original image as a template image; Using the rectangular positioning frame to intercept a corresponding local image from the video stream, performing a differential calculation between the local image and the template image to obtain a differential image, counting the motion area of ​​each frame according to the differential image, and filtering abnormal differential images according to preset rules; The preset rule includes: counting the proportion of the motion area, and if the proportion is greater than a preset proportion, repeating the preprocessing operation, and updating the rectangular positioning frame and the template image; Presetting a pixel deviation threshold, binarizing the difference image using the pixel deviation threshold, binarizing the filtered difference image, and then removing interference contours in the binary image; According to the prior conditions, the pointer contour that meets the preset conditions is selected from the binary image as the final pointer contour; the selecting the pointer contour that meets the preset conditions from the binary image as the final pointer contour according to the prior conditions includes: setting a confidence threshold, setting the prior conditions according to the pointer characteristics, and calculating the confidence of each pointer contour in the binary image according to the prior conditions, if the confidence corresponding to all the pointer contours is less than the confidence threshold, then attenuating the pixel deviation threshold with a preset step size, and returning to the step of removing the interfering contour in the binary image until multiple candidate contours are obtained, and taking the pointer contour with the highest confidence among the candidate contours as the final pointer contour; the prior conditions include pointer shape, length and / or orientation; The pointer endpoint coordinates of the final pointer outline in each frame of the binary image are recorded, and it is determined whether the change value of the pointer endpoint coordinates within a preset time window is stable within a preset range. If so, a balance determination is performed.

2. The method for judging the balance state of a pan balance pointer for physical, chemical and biological experiments according to claim 1, characterized in that: The preprocessing includes: performing white threshold segmentation on the target image to obtain a white pallet outline and a white dial outline, relocating the original color dial outline in the target image according to the relative position relationship between the white pallet outline and the white dial outline, using the minimum outer rectangle of the original color dial outline and expanding the rectangle according to the shape ratio of the dial to obtain the rectangular positioning frame including the dial and the pointer, and overlaying the rectangular positioning frame on the original image as the template image.

3. The method for judging the balance state of a pan balance pointer for physical, chemical and biological experiments according to claim 1, characterized in that: The balance determination includes: presetting a scale plate area as a balance interval of the balance, and if the coordinates of the pointer endpoint fall into the balance interval, it is determined that the balance is balanced at the current moment.

4. A system for judging the balance state of a pan balance pointer for physical, chemical and biological experiments, characterized in that: include: The original image acquisition module is used to intercept the original image of the experimenter adjusting the balance of the tray from the video stream, and intercept the target image containing the tray from the original image; A template image generation module is used to pre-process the target image, obtain a rectangular positioning frame including a dial and a pointer, and overlay the rectangular positioning frame onto the original image as a template image; An abnormal image filtering module is used to intercept a corresponding local image from the video stream using the rectangular positioning frame, perform a differential calculation between the local image and the template image to obtain a differential image, count the motion area of ​​each frame according to the differential image, and filter the abnormal differential image according to a preset rule; The preset rule includes: counting the proportion of the motion area, and if the proportion is greater than a preset proportion, repeating the preprocessing operation, and updating the rectangular positioning frame and the template image; A noise filtering module, which presets a pixel deviation threshold, uses the pixel deviation threshold to binarize the differential image, and is used to binarize the filtered differential image and then remove the interfering contour in the binary image; A pointer contour recognition module, used for selecting a pointer contour meeting a preset condition from the binary image according to a priori conditions as a final pointer contour; the selecting a pointer contour meeting a preset condition from the binary image according to the priori conditions as the final pointer contour comprises: setting a confidence threshold, setting the priori conditions according to pointer features, and calculating the confidence of each pointer contour in the binary image according to the priori conditions, if the confidence corresponding to all the pointer contours is less than the confidence threshold, then attenuating the pixel deviation threshold with a preset step size, and returning to the step of removing the interfering contour in the binary image until a plurality of candidate contours are obtained, and taking the pointer contour with the highest confidence among the candidate contours as the final pointer contour; the priori conditions include pointer shape, length and / or orientation; The pointer balance recognition module is used to record the pointer endpoint coordinates of the final pointer outline in each frame of the binary image, and determine whether the change value of the pointer endpoint coordinates in a preset time window is stable within a preset range. If so, a balance determination is performed.

5. A device for judging the balance state of a tray balance pointer for physical, chemical and biological experiments, characterized in that: The device includes a memory, a processor and a communication module, wherein: The memory is used to store executable program code; The processor is coupled to the memory; The processor calls the executable program code stored in the memory to execute the method for determining the balance state of a pan balance pointer for physical, chemical and biological experiments as described in any one of claims 1 to 3.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and when the computer instructions are called, they are used to execute the method for judging the balance state of a tray balance pointer for physical, chemical and biological experiments as described in any one of claims 1 to 3.

Images