A method and system for automatically scoring a physical-chemical test based on a drop event

Abstract

The application provides a kind of automatic scoring method and system of physicochemical test based on drop event, comprising: using mixed Gaussian model to model the change of gray value of the region to be tested in the experiment process;Using the established mixed Gaussian model, the gray value change of the region to be tested in the video to be tested is judged, and whether a drop event occurs is determined according to the judgment result;According to the number of the drop event, automatically judge whether the physicochemical test is scored.The application uses mixed Gaussian model, the computing power cost is low, and only CPU can complete the calculation;The calculation speed is fast, and it can reach more than 50FPS on the CPU of intel i7-8700;The accuracy is as high as 85%.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of machine vision, in particular to a method and system for automatically scoring a physical and chemical test based on a droplet event. BACKGROUND

[0002] In recent years, the number of examinees for the national college entrance examination has been increasing, which has posed a high challenge to the demand for human resources in the examination and evaluation process. Among them, the physical and chemical test examination has shown the phenomenon of fewer teachers and more examinees due to its multiple scoring points, complex process, and long evaluation time, and the evaluation quality has been greatly tested.

[0003] Recently, in order to reduce the human cost of evaluation in the physical and chemical experiment examination, an automatic scoring system for physical and chemical experiment examination based on machine vision was proposed. This method uses multiple cameras to collect image information of examinees doing experiments, and then scores the scoring points through multiple algorithm modules in the back end. Among them, there is an evaluation of a scoring point in the "filtration experiment", which is to judge whether the lower end of the sharp mouth funnel needs to be close to the beaker wall in the filtration process to prevent droplets from splashing. Close to the beaker wall is scored, otherwise it is not scored, see Figure 1 .

[0004] After searching, the Chinese patent with application number 202010203774.X discloses a chemical funnel device building experiment operation device and intelligent scoring method. It uses Hough circle detection and non-maximum suppression algorithm to extract the approximate circles of the long-neck funnel and the funnel mouth and beaker mouth in the overhead image, and according to the distance between the centers of the two circles in the image, the height difference information of the two object circle mouths, and the internal parameters of the overhead camera, the accurate distance between the centers of the two circles is obtained, and according to the distance and the radius of the beaker mouth, it is judged whether the lower end of the funnel is close to the inner wall of the beaker. This patent has the following problems: 1) In the overhead view, the process of the sharp mouth funnel blocking the beaker often occurs, and the extraction of the beaker center is prone to deviation; 2) The height of the sharp mouth funnel is uncertain in each experiment, and the changing height will cause inaccurate distance calculation of the circle centers; 3) It cannot be assumed that the lower end of the sharp mouth funnel and the center of the sharp mouth funnel coincide in the vertical direction, and the sharp mouth funnel may be slightly tilted; 4) Close and not close often only have a difference of 1mm. Due to the above problems, the accuracy of this method cannot meet the demand, so it cannot effectively judge whether the lower end of the sharp mouth funnel is close to the beaker wall.

[0005] Through retrieval, the Chinese patent with application number 202210898308.7 discloses a method for determining whether a sharp funnel is tightly close to the wall of a beaker through drop detection. The method considers that, in a "filtration experiment", if the lower end of the sharp funnel is not tightly close to the wall of the beaker, the filtered liquid will directly drip on the bottom of the beaker wall along the lower end of the sharp funnel (hereinafter referred to as a drop event). By judging whether a drop event exists, it can be determined whether the lower end of the sharp funnel is tightly close to the inner wall of the beaker. As for the specific method of detecting the drop event, the patent first extracts the optical flow map of each frame through an optical flow model, and then extracts the picture and optical flow of each frame in the video Figure 1 into a sequence model to determine the number of drop events occurring during the experiment. If the number is greater than a certain threshold, it is considered that the lower end of the sharp funnel is not tightly close to the wall of the beaker. The problems of this method are: 1) slow processing speed, the processing speed of the optical flow extraction model and the sequence model is much lower than real-time; 2) high cost of computing power, two GPU cards (1080ti or above) are required for simultaneous use of the two models; 3) the optical flow extraction model cannot well capture the motion trajectory of the water droplet, the shape of the liquid droplet is not fixed and is prone to blur during falling, which poses challenges to the extraction and matching of features in the optical flow extraction process. SUMMARY

[0006] In view of the defects in the prior art, the purpose of the present application is to provide a physicochemical test automatic scoring method and system based on drop events.

[0007] According to one aspect of the present application, a physicochemical test automatic scoring method based on drop events is provided, comprising:

[0008] using a mixture Gaussian model to model the change of the gray value of the to-be-tested region during the experiment;

[0009] using the established mixture Gaussian model to judge the change of the gray value of the to-be-tested region in the to-be-tested video, and determining whether a drop event has occurred according to the judgment result;

[0010] automatically judging whether the physicochemical test is scored according to the number of drop events.

[0011] Preferably, the to-be-tested region is the union region of the funnel and the beaker.

[0012] Preferably, the judgment of the change of the gray value of the to-be-tested region comprises:

[0013] judging the distance between the gray value of the to-be-tested region and the gray value distribution of the region, and if the distance exceeds a first threshold, the corresponding gray value is judged as foreground;

[0014] if the number of foreground pixels exceeds a second threshold, it is determined that a drop event has occurred.

[0015] Preferably, the step of determining that a dripping event has occurred when the number of pixels in the foreground exceeds a third threshold includes:

[0016] The distance between the gray value of each pixel and the Gaussian mixture distribution of the gray values ​​of that pixel is represented by a heatmap.

[0017] When the area of ​​pixels with a heat value greater than the first threshold in the heatmap is greater than the second threshold, a dripping event is considered to have occurred.

[0018] Preferably, the step of automatically determining whether a score has been awarded for the physicochemical test based on the number of droplet events includes:

[0019] If the number of dripping events during the experiment exceeds the third threshold, the experiment is considered to have a dripping event and no points are awarded; otherwise, points are awarded.

[0020] Preferably, the video to be tested is obtained by shooting the experiment using a head-up camera.

[0021] According to a second aspect of the present invention, an automatic scoring system for physicochemical tests based on droplet events is provided, comprising:

[0022] The Gaussian mixture modeling module uses a Gaussian mixture model to model the changes in grayscale values ​​of the test area during the experiment.

[0023] The dripping event detection module uses a pre-established Gaussian mixture model to judge the changes in grayscale values ​​of the test area in the test video, and determines whether a dripping event has occurred based on the judgment result.

[0024] The calculation module automatically determines whether the physicochemical test scores based on the number of droplet events.

[0025] According to a third aspect of the present invention, an automatic scoring device for physicochemical tests based on a droplet event is provided, employing any of the methods described above, or employing the system described above.

[0026] According to a fourth aspect of the present invention, a terminal is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, can be used to perform the methods described above, or to run the systems described above.

[0027] According to a fifth aspect of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, can be used to perform the methods described above, or to run the system described above.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] The automatic scoring method and system for physicochemical tests based on droplet events in this invention uses a Gaussian mixture model, has low computational cost, and can complete the calculation with only a CPU.

[0030] The automatic scoring method and system for physicochemical tests based on droplet events in this invention uses a Gaussian mixture model, which has a fast calculation speed and can reach more than 50 FPS on an Intel i7-8700 CPU.

[0031] The automatic scoring method and system for physicochemical tests based on droplet events in this invention embodiment has an accuracy rate of up to 85%. Attached Figure Description

[0032] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0033] Figure 1 This is a schematic diagram illustrating the scoring criteria for whether the lower end of the pointed funnel is in close contact with the beaker wall during a physicochemical experiment.

[0034] Figure 2 This is a flowchart of an automatic scoring method for physicochemical tests based on droplet events, according to one embodiment of the present invention.

[0035] Figure 3 This is a thermogram of a droplet under a mixed Gaussian background modeling in a preferred embodiment of the present invention.

[0036] Figure 4 This is a flowchart of an automatic scoring method for physicochemical tests based on droplet events according to a preferred embodiment of the present invention. Detailed Implementation

[0037] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention. These all fall within the scope of protection of the present invention.

[0038] See Figure 2 This invention provides an embodiment of an automatic scoring method for physicochemical tests based on droplet events, comprising:

[0039] S100, using a Gaussian mixture model to model the changes in grayscale values ​​of the test area during the experiment;

[0040] S200 uses a pre-established Gaussian mixture model to judge the grayscale value changes in the test area of ​​the test video, and determines whether a dripping event has occurred based on the judgment result;

[0041] S300 automatically determines whether a physicochemical test is scored based on the number of dripping events.

[0042] In this embodiment, a Gaussian mixture model is used, which has low computational cost and can complete the calculation with only a CPU.

[0043] In a preferred embodiment of the present invention, a method for obtaining the test area in S100 is provided. Specifically, firstly, an experimental video is obtained; then, the position of the beaker is obtained through the target detection module, and the beaker area is extracted as the test area; finally, the test area is grayscale processed.

[0044] In a preferred embodiment, the area to be measured is the union of the areas where the funnel and the beaker are located, i.e., the area where the droplet falls and the area where the droplet ripples are observed.

[0045] In a preferred embodiment of the present invention, implementing S200 specifically includes the following steps:

[0046] S201, determine the distance between the gray value of the area to be measured and the gray value distribution. If it exceeds the first threshold, the corresponding gray value is judged as the foreground.

[0047] S202, when the number of foreground pixels determined in S201 exceeds the second threshold, a dripping event is identified.

[0048] In a preferred embodiment, step S202 is implemented, specifically, the distance between the grayscale value of each pixel and the Gaussian distribution of that pixel's grayscale value is represented by a heatmap, as follows: Figure 3 As shown, the thermal values ​​of the dripping area and the water ripple area generated by the dripping are significantly higher than those of other areas. The dynamic area can be easily distinguished by the first thermal threshold. If the area of ​​the moving area is greater than the second threshold, it is considered that a dripping event has occurred.

[0049] The method for identifying dripping events in this embodiment is simple and fast to calculate, and it hardly occupies any computing memory.

[0050] In a preferred embodiment of the present invention, in step S300, if the number of dripping events during the experiment exceeds a third threshold, the experiment is deemed to have a dripping event and is judged to receive no score; otherwise, the experiment is scored.

[0051] Based on the same inventive concept, other embodiments of the present invention provide an automatic scoring system for physicochemical tests based on dripping events, including a Gaussian mixture modeling module, a dripping event detection module, and a calculation module; the Gaussian mixture modeling module uses a Gaussian mixture model to model the changes in grayscale values ​​of the test area during the experiment; the dripping event detection module uses the established Gaussian mixture model to judge the changes in grayscale values ​​of the test area in the test video, and determines whether a dripping event has occurred based on the judgment result; the calculation module automatically determines whether the physicochemical test is scored based on the number of dripping events.

[0052] In the dripping event detection module, the grayscale value distribution of each pixel in the test area is statistically analyzed in each frame of the test video. For each frame of the test video, the distance between the grayscale value of each pixel in the test area and the grayscale value distribution of that pixel is calculated. If the distance exceeds the first threshold, the corresponding pixel is identified as the foreground. When the number of pixels in the foreground exceeds the second threshold, a dripping event is identified as having occurred in that frame.

[0053] In the calculation module, if the number of dripping events during the experiment exceeds the third threshold, the experiment is considered to have a dripping event and no points are awarded; otherwise, points are awarded.

[0054] Based on the same inventive concept, other embodiments of the present invention provide an automatic scoring device for physicochemical tests based on droplet events, employing the above-described method or system.

[0055] Based on the same inventive concept, other embodiments of the present invention provide a terminal, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it can be used to perform the above-described method or run the above-described system.

[0056] Based on the same inventive concept, other embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, can be used to perform the above-described method or to run the above-described system.

[0057] See Figure 4 In a preferred embodiment of the present invention, a complete operation process is provided, specifically including:

[0058] Step 1: Obtain the experimental video.

[0059] Step 2: Use the target detection module to obtain the position of the beaker and extract the beaker area (i.e., the target area or the area to be tested).

[0060] Step 3: Grayscale conversion of the target area.

[0061] Step 4: Locate the motion region of the target area using Gaussian mixture background modeling. During the dripping process, the grayscale values ​​of pixels in the dynamic target area, such as the dripping liquid and the ripples it generates, change significantly and are far from the Gaussian mixture distribution of the background pixels. If we represent the grayscale value of each pixel and its distance from the Gaussian mixture distribution using a heatmap, then... Figure 3 As shown, the thermal values ​​of the dripping area and the water ripple area generated by the dripping are significantly higher than those of other areas, and this dynamic area can be easily distinguished by the first thermal threshold.

[0062] Step 5: If the area of ​​the moving region is greater than the second threshold, then a dripping event is considered to have occurred.

[0063] Step 6: If the number of dripping events is greater than the third threshold during the target time period of the video, it is considered that the lower end of the pointed funnel is not in close contact with the inner wall of the beaker; otherwise, it is considered that the lower end of the pointed funnel is in close contact with the inner wall of the beaker.

[0064] This embodiment was applied to 100 experimental videos (50 videos with operation scores and 50 videos with operation scores), and the judgment was made using the method of this embodiment. The judgment accuracy was acc≥85%.

[0065] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0066] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0067] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0068] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0069] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. An automatic scoring method for physicochemical tests based on droplet events, characterized in that, include: A Gaussian mixture model was used to model the changes in grayscale values ​​of the test area during the experiment. Using the established Gaussian mixture model, the grayscale value changes of the test area in the test video are judged, and the judgment results are used to determine whether a dripping event has occurred. The system automatically determines whether the physicochemical test scores based on the number of droplet events. The judgment of grayscale value changes in the area to be tested includes: The grayscale value distribution of each pixel in the test area in each frame of the test video is statistically analyzed. For each frame in the video to be tested, calculate the distance between the gray value of each pixel in the test area and the gray value distribution of that pixel. If the distance exceeds the first threshold, the corresponding pixel is judged as the foreground. When the number of pixels in the foreground of a frame exceeds the second threshold, it is determined that a dripping event has occurred in that frame. When the number of foreground pixels in a frame exceeds a second threshold, it is determined that a dripping event has occurred in that frame, including: Calculate the distance between the gray value of each pixel and the mixture Gaussian distribution of the gray values ​​of that pixel, and represent it using a heatmap; When the area in the heat map where the thermal value is greater than the first threshold is greater than the second threshold, a dripping event is considered to have occurred. The automatic determination of whether a score is given for a physicochemical test based on the number of droplet events includes: If the number of dripping events during the experiment exceeds the third threshold, the experiment is considered to have a dripping event and no points are awarded; otherwise, points are awarded.

2. The automatic scoring method for physicochemical tests based on droplet events according to claim 1, characterized in that, The area to be tested is the union of the areas where the funnel and the beaker are located.

3. The automatic scoring method for physicochemical tests based on droplet events according to claim 1, characterized in that, The video to be tested was obtained during the experiment using a head-up camera.

4. An automatic scoring system for physicochemical tests based on droplet events, characterized in that, include: The Gaussian mixture modeling module uses a Gaussian mixture model to model the changes in grayscale values ​​of the test area during the experiment. The dripping event detection module uses a pre-established Gaussian mixture model to judge the changes in grayscale values ​​of the test area in the test video, and determines whether a dripping event has occurred based on the judgment result. The calculation module automatically determines whether the physicochemical test is scored based on the number of droplet events. The judgment of grayscale value changes in the area to be tested includes: The grayscale value distribution of each pixel in the test area in each frame of the test video is statistically analyzed. For each frame in the video to be tested, calculate the distance between the gray value of each pixel in the test area and the gray value distribution of that pixel. If the distance exceeds the first threshold, the corresponding pixel is judged as the foreground. If the number of pixels in the foreground of a frame exceeds the second threshold, then a dripping event is considered to have occurred in that frame. When the number of foreground pixels in a frame exceeds a second threshold, it is determined that a dripping event has occurred in that frame, including: Calculate the distance between the gray value of each pixel and the mixture Gaussian distribution of the gray values ​​of that pixel, and represent it using a heatmap; When the area in the heat map where the thermal value is greater than the first threshold is greater than the second threshold, a dripping event is considered to have occurred.

5. A terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it can be used to perform the method of any one of claims 1-3, or to run the system of claim 4.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program can be used to perform the method of any one of claims 1-3, or to run the system of claim 4.

Images