Intelligent control method and system for hot dry noodle shaking equipment

Abstract

The present invention relates to the field of data processing, and specifically to an intelligent control method and system for a hot dry noodle shaking device, the method comprising: preprocessing a hot dry noodle image on a historical shaking device to obtain a historical grayscale image; performing edge detection on the historical grayscale image to obtain a historical edge image, calculating the uniformity of edge distribution in the historical edge image, normalizing the uniformity of edge distribution to obtain the shaking degree of the historical hot dry noodle image; collecting a current hot dry noodle image and calculating the corresponding shaking degree, taking the ratio of the shaking degree of the current hot dry noodle image to the mean of the shaking degree of the historical hot dry noodle image as a power control coefficient of the current shaking device, multiplying the power control coefficient by the vibration intensity of the current shaking device to obtain an adjusted vibration intensity, and shaking the hot dry noodles based on the adjusted vibration intensity. The present invention realizes precise control and optimization of the hot dry noodle shaking process, and improves the uniformity of the shaking effect.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and more specifically, to an intelligent control method and system for hot dry noodle shaking equipment. Background Art

[0002] Hot dry noodles is a traditional delicacy. With the expansion of the consumer market, its production demand is also increasing year by year. In the production process of hot dry noodles, in order to prevent the noodles from sticking together, the hot dry noodles need to be shaken out by a shaking device. The judgment of the degree of shaking out traditionally relies on manual labor, which is time-consuming and labor-intensive. Therefore, it is necessary to find or develop a more accurate and objective method for judging the degree of shaking out, so as to better control the shaking out device and improve production efficiency and product quality.

[0003] Prior art, such as the patent application document with publication number CN108782797A, discloses a control method for frying tea leaves in an arm-type tea frying machine and an arm-type tea frying machine. The control method for frying tea leaves in the arm-type tea frying machine includes: real-time shooting of tea leaves in a circular frying pan during the tea frying process, obtaining a first image of the initial frying of the upper surface of the tea leaves and a second image after frying for a predetermined time; determining that the distribution shape of the tea leaves in the second image is not circular, performing ROI superposition processing on the first image and the second image to obtain a third image; performing Canny edge detection processing on the third image to obtain the coordinates of the edge points, and then calculating the rotation matrix based on the coordinates of the edge points; controlling the motor to rotate the left and right claws according to the rotation matrix and then frying the tea leaves. The invention automatically adjusts the frying angle of the tea frying robot through real-time shooting with an industrial camera and machine vision technology, ensuring that the tea leaves are evenly distributed in the pot, and realizing an efficient and uniform withering process.

[0004] However, the edge detection of the image in the above operation does not take into account the stacking of tea leaves, so that the edge detection result cannot accurately reflect the real edge, resulting in a decrease in subsequent control efficiency. Summary of the invention

[0005] In order to solve the technical problem that the above-mentioned edge detection result cannot accurately reflect the real edge, resulting in a decrease in subsequent control efficiency, the present invention provides solutions in the following aspects.

[0006] In a first aspect, an intelligent control method for a hot dry noodle shaking device comprises: Collect multiple hot-dry noodles images from historical shaking devices, and pre-process the images to obtain historical grayscale images; Perform edge detection on the historical grayscale image to obtain a historical edge image, calculate the uniformity of edge distribution in the historical edge image, normalize the uniformity of edge distribution to obtain the degree of jitter of the historical hot-dry noodle image; calculate the mean value of the degree of jitter of all historical hot-dry noodle images; Collect the current hot-dry noodles image and calculate the corresponding shaking degree, take the ratio of the shaking degree of the current hot-dry noodles image to the average shaking degree of the historical hot-dry noodles images as the power control coefficient of the current shaking device, multiply the power control coefficient by the vibration intensity of the current shaking device to obtain the adjusted vibration intensity, and shake the hot-dry noodles based on the adjusted vibration intensity.

[0007] The present invention performs edge detection on historical hot-dry noodles grayscale images, takes into account the situation that hot-dry noodles are stacked during the shaking process of hot-dry noodles, calculates the uniformity of edge distribution based on the edge detection result, and uses the uniformity of edge distribution as an indicator for evaluating image quality. For historical shaking devices, the calculation result of the uniformity of edge distribution is fed back to the control system, and the edge distribution of the current hot-dry noodles image is further collected and analyzed in real time, compared with historical data, and the vibration intensity is adjusted according to the ratio to achieve dynamic adjustment, thereby ensuring that each shaking process can achieve the best effect.

[0008] Preferably, the performing edge detection on the historical grayscale image to obtain the historical edge image further comprises: Obtain all edge connected domains in the edge image and count the number of pixels contained in each connected domain; For each connected domain, Freeman chain code is used to represent its edge contour, and a chain code sequence is constructed, where each chain code value represents the direction from the current pixel to the next pixel; Calculates the chain code increment for each chain code value in the chain code sequence.

[0009] Freeman chain codes are used to represent edge contours. Each chain code represents the direction from the current pixel to the next pixel. This representation method can accurately describe the shape and structure of the edge and calculate the increments of each chain code value in the chain code sequence. These increments reflect the complexity and degree of change of the edge. By analyzing the chain code increments, we can better understand the distribution of edges in the image, thereby optimizing the operating parameters of the jitter device.

[0010] Preferably, the process of obtaining the distribution uniformity includes: The stretch degree of each connected domain is calculated, and the mean of the stretch degrees of all connected domains is used as the distribution uniformity; the stretch degree satisfies the relationship: ; In the formula, For the The stretch degree of the connected domain, For the The standard deviation of the chain code increments for all chain code values ​​in a connected domain, For the The total number of pixels contained in a connected domain, For the The number of chain codes in connected domains, For the The first The chaincode increment for each chaincode.

[0011] The calculation of the stretch degree involves the increment of all chain codes in the connected domain, which helps to identify the complete outline of the edge, avoid edge breakage or missing problems caused by local noise or poor image quality, and make the detected edge more complete and continuous.

[0012] Preferably, after obtaining the distribution uniformity, the method further includes optimizing the distribution uniformity to obtain an optimized distribution uniformity, wherein the optimized distribution uniformity satisfies the following relationship: ; In the formula, is the uniformity of the distribution after optimization, For the The number of centers whose distances from the centers of connected domains are less than the distribution radius. is the total number of connected domains, For the The total number of pixels contained in a connected domain, is the average center point density under uniform distribution, represents the natural exponential function.

[0013] The optimized distribution uniformity formula introduces factors such as the natural exponential function and the distance between the center points of the connected domain, which can more comprehensively consider the distribution characteristics of hot-dry noodles in the image. The natural exponential function has the characteristics of smoothness and attenuation, which can effectively suppress the extreme values ​​of the distribution uniformity and reduce the evaluation deviation caused by local anomalies, thereby improving the accuracy of the jitter degree evaluation.

[0014] Preferably, the average center point density under uniform distribution The relationship is satisfied: ; In the formula, is the total number of connected domains, is the total number of connected domains, is the number of all pixels contained in the edge image, is the distribution radius; the average value of the number of pixels contained in all connected domains is taken as the distribution radius.

[0015] The average center point density under uniform distribution is a standardized reference value that reflects the average number of center points per unit area under uniform distribution. By comparing it with the actual center point density, the degree of dispersion of hot-dry noodles can be more accurately evaluated. When the actual center point density is close to the average center point density under uniform distribution, it means that the distribution of hot-dry noodles is relatively uniform.

[0016] Preferably, the degree of jitter of the historical hot-dry noodle image satisfies the relationship: ; In the formula, The degree of looseness of the hot dry noodles. is the uniformity of edge distribution, is the maximum value of the uniformity of edge distribution, is the minimum value of edge distribution uniformity.

[0017] Preferably, the edge detection adopts the Canny edge detection algorithm.

[0018] In a second aspect, an intelligent control system for a hot dry noodle shaking device includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the intelligent control method for the hot dry noodle shaking device is implemented.

[0019] The beneficial effects of the present invention are: By collecting historical and current hot-dry noodles images, preprocessing and edge detection of the images, and calculating the uniformity of edge distribution and its normalized value, the shaking effect of hot-dry noodles can be accurately evaluated.

[0020] According to the ratio of the current degree of dispersion of the hot-dry noodles image to the historical average, the power control coefficient and vibration intensity of the dispersion device are dynamically adjusted. This adaptive adjustment mechanism can make the dispersion effect of the hot-dry noodles more uniform and avoid the problem of local aggregation or uneven dispersion. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1 It is a method flow chart of steps S1 to S3 in an intelligent control method of a hot dry noodle shaking device according to an embodiment of the present invention.

[0022] Figure 2 This is an image of hot-dry noodles taken from a shaking device.

[0023] Figure 3 It is a grayscale image of a sample hot-dry noodles image captured from the hot-dry noodles image.

[0024] Figure 4 It is the edge image obtained by edge detection of grayscale image.

[0025] Figure 5 It is an image of the distinct edges of hot-dry noodles after they are shaken apart.

[0026] Figure 6 It is an image of the edges of hot-dried noodles tangled together in a messy way. DETAILED DESCRIPTION

[0027] 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0028] The embodiment of the present invention discloses an intelligent control method for a hot dry noodle shaking device, referring to Figure 1 , including steps S1 to S3, which are specifically as follows: S1: Collect multiple hot-dry noodles images from historical shaking devices and pre-process the images to obtain historical grayscale images.

[0029] A high-definition camera is installed on the shaking device to capture the image of the hot-dried noodles in real time during the shaking process. Figure 2 The figure shows the overall state of hot dry noodles on the shaking device. Then, multiple historical hot dry noodles images were collected.

[0030] Furthermore, taking a single historical image as an example, a portion of the collected overall image (such as Figure 2 The image shown in the black box is used as the sample hot-dry noodles image. This part of the captured image can be marked as the hot-dry noodles image and used as a basis for judging the current degree of hot-dry noodles shaking.

[0031] The intercepted hot dry noodles image is grayed out to obtain a historical grayscale image (such as Figure 3 as shown).

[0032] S2: Perform edge detection on the historical grayscale image to obtain a historical edge image, calculate the uniformity of edge distribution in the historical edge image, normalize the uniformity of edge distribution to obtain the degree of jitter of the historical hot-dry noodles image; calculate the mean of the degree of jitter of all historical hot-dry noodles images.

[0033] Use edge detection algorithm to perform edge detection on the historical grayscale image of S1 (Canny edge detection algorithm, Sobel operator, etc. can be used) to obtain the historical edge image (such as Figure 4 As shown in the figure), obtain all edge connected domains in the historical edge image and count the number of pixels contained in each connected domain.

[0034] In addition, the edge images of the other two types of hot dry noodles are shown. Figure 5 and Figure 6 , Figure 5This is the image of the distinct edges of the hot dry noodles after they are shaken apart. Figure 6 The image of the edges of hot-dried noodles are tangled together in a messy way.

[0035] For each connected domain, the Freeman chain code is used to represent its edge contour, and a chain code sequence is constructed, where each chain code value represents the direction from the current pixel to the next pixel.

[0036] Among them, the Freeman chain code is a compact edge representation method that reduces the amount of data by decomposing the edge contour into a series of direction vectors. Each direction vector is represented by a specific numerical value, which corresponds to one of a fixed set of directions. In the most common 8-direction chain code, these directions usually correspond to integers from 0 to 7, each of which represents a specific angle range (for example, 0 may represent a 0-degree direction, 1 represents a 45-degree direction, and so on, until 7 represents a 315-degree direction).

[0037] Further calculation of the chain code increment of each chain code value, it should be noted that in the Freeman chain code, the chain code value is usually a cyclic sequence, because the direction is periodic. When calculating the chain code increment, it is actually calculating the "minimum direction change" from one direction to another.

[0038] Therefore, when changing from a higher chain code value (such as 7) to a lower chain code value (such as 0 or 1), the cyclic nature of the chain code value needs to be taken into account.

[0039] For example, the second chain code value is set to 3 and the third chain code value is set to 5. Then the chain code value changes from 3 to 5, and the direct direction change is +2 (that is, it rotates clockwise for 2 direction intervals). Therefore, the chain code increment of the second chain code value is 2.

[0040] For example, the fourth chain code value is then set to 1, and the previous (third) chain code value is 5. If the difference between 5 and 1 is directly calculated, the chain code increment is -4, but this does not make sense in the direction change because the direction is cyclic. Therefore, it is necessary to find the minimum direction change from 5 to 1. This can be achieved by continuously adding 1 (and considering the cycle) until 1 is reached.

[0041] So, starting from 5: 5+1=6, 6+1=7, 7+1=0 (loops to 0), 0+1=1.

[0042] Therefore, the chain code increment of the third chain code value is 4 (indicating 4 clockwise rotation intervals, taking into account the cyclicity).

[0043] The calculation of the chain code increments for the remaining chain code values ​​can be obtained by following the same loop logic as above.

[0044] The stretch degree of the edge connected domain is further calculated by analyzing the edge trend of each connected domain, the size of the edge connected domain, etc., which reflects the smoothness, continuity and uniformity of the edge. A higher stretch degree means that the edge is smoother, more continuous and relatively evenly distributed.

[0045] Then the stretch degree of the connected domain satisfies the relationship:

[0046] In the formula, For the The stretch degree of the connected domain, For the The standard deviation of the chain code increments for all chain code values ​​in a connected domain, For the The total number of pixels contained in a connected domain, For the The number of chain codes in connected domains, For the The first The chaincode increment for each chaincode.

[0047] in, It is used to measure the discreteness of the chain code increment, that is, the volatility of the marginal trend. When it is small, it means that the chain code increment is relatively concentrated and the marginal trend is relatively stable; otherwise, it means that the chain code increment is dispersed and the marginal trend is relatively volatile.

[0048] The setting is the above The result of normalization is that the maximum value of the chain code increment is 7 (for example, in the case of eight-connectivity, a pixel can be connected to at most 7 pixels around it).

[0049] It means summing up the deviations of all chain code increments to get the total deviation of chain code increments in the entire edge connected domain. This total deviation reflects the overall volatility of the edge trend.

[0050] According to the above calculation The stretch degree of each connected domain can be calculated in the same way as the stretch degree of each edge connected domain. Specifically, if the stretch degree of each edge connected domain is high, it means that the edge is relatively smooth, continuous, and the distribution is relatively uniform. On the contrary, if the stretch degree of some edge connected domains is low, it may mean that these edges are broken, irregular, or over-aggregated, thereby reducing the uniformity of edge distribution.

[0051] Then, the mean of the stretching degrees of all connected domains is taken as the uniformity of edge distribution in the edge image.

[0052] It should be noted that the spatial distribution of the connected domain needs to be considered to further optimize the uniformity of the edge distribution calculated above.

[0053] Specifically, the coordinates of the center point of each connected domain are calculated, and the average number of pixels contained in all connected domains is taken as the distribution radius. The number of center points whose distance from the center point of any connected domain is less than the distribution radius is further counted, and then the uniformity of the edge distribution calculated above is optimized. The optimized uniformity of the distribution satisfies the relationship:

[0054] In the formula, is the uniformity of the distribution after optimization, For the The number of centers whose distances from the centers of connected domains are less than the distribution radius. is the total number of connected domains, For the The total number of pixels contained in a connected domain, is the average center point density under uniform distribution, represents the natural exponential function.

[0055] In the formula The part represents the weighted average number index of the center points within the distribution radius of all the center points of the edge connected domain. The more edge pixels an edge connected domain contains, the larger its influence range is considered to be. The number of other center points within its distribution radius can be appropriately increased relative to other center points, and the weighted average value is calculated accordingly.

[0056] Among them, the average center point density under uniform distribution reflects the idealized average number of center points around each center point under the assumption of uniform distribution. The average center point density under uniform distribution is The relationship is satisfied:

[0057] In the formula, is the total number of connected domains, is the total number of connected domains, is the number of all pixels contained in the edge image, is the distribution radius.

[0058] The calculated edge distribution uniformity takes into account both the stretch degree of each connected domain in the image and the distribution of the connected domains.

[0059] Therefore, the degree of shaking of hot-dry noodles is obtained by normalizing the uniformity of edge distribution.

[0060] The degree of shaking of hot dry noodles satisfies the relationship:

[0061] In the formula, The degree of looseness of the hot dry noodles. is the uniformity of edge distribution, is the maximum value of the uniformity of edge distribution, is the minimum value of edge distribution uniformity.

[0062] According to the above method for calculating the shaking degree of hot-dry noodles of a single image, the shaking degree of hot-dry noodles corresponding to multiple collected historical images can be calculated in the same way.

[0063] S3: Collect the current hot-dry noodles image and calculate the corresponding shaking degree, take the ratio of the shaking degree of the current hot-dry noodles image to the average shaking degree of the historical hot-dry noodles images as the power control coefficient of the current shaking device, multiply the power control coefficient by the vibration intensity of the current shaking device to obtain the adjusted vibration intensity, and shake the hot-dry noodles based on the adjusted vibration intensity.

[0064] Calculate the average value of the degree of dispersion of all historical images collected, and obtain the latest collected hot-dry noodles image, calculate its degree of dispersion according to the calculation method of S1-S2 above, compare the degree of dispersion of the latest collected image with the average value of the degree of dispersion of all historical images, use the ratio as the power control coefficient of the dispersion device, multiply the control coefficient by the vibration intensity of the current dispersion device to obtain the adjusted vibration intensity, and disperse the hot-dry noodles based on the adjusted vibration intensity.

[0065] An embodiment of the present invention also discloses an intelligent control system for a hot dry noodle shaking device, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an intelligent control method for the hot dry noodle shaking device according to the present invention is implemented.

[0066] The system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface. The configuration and functions of these components are known in the art and will not be described in detail here.

[0067] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module or both. Any such computer storage medium may be part of a device or accessible or connectable to a device.

[0068] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0069] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. An intelligent control method for hot dry noodle shaking equipment, characterized in that: include: Collect multiple hot-dry noodles images from historical shaking devices, and pre-process the images to obtain historical grayscale images; Perform edge detection on the historical grayscale image to obtain a historical edge image, calculate the uniformity of edge distribution in the historical edge image, and normalize the uniformity of edge distribution to obtain the degree of jitter of the historical hot-dry noodles image; Calculate the average of the degree of jitter of all historical hot-dry noodles images; Collect the current hot-dry noodles image and calculate the corresponding shaking degree, take the ratio of the shaking degree of the current hot-dry noodles image to the average shaking degree of the historical hot-dry noodles images as the power control coefficient of the current shaking device, multiply the power control coefficient by the vibration intensity of the current shaking device to obtain the adjusted vibration intensity, and shake the hot-dry noodles based on the adjusted vibration intensity.

2. The intelligent control method for hot dry noodle shaking equipment according to claim 1 is characterized in that: The performing edge detection on the historical grayscale image to obtain the historical edge image further comprises: Obtain all edge connected domains in the edge image and count the number of pixels contained in each connected domain; For each connected domain, Freeman chain code is used to represent its edge contour, and a chain code sequence is constructed, where each chain code value represents the direction from the current pixel to the next pixel; Calculates the chain code increment for each chain code value in the chain code sequence.

3. The intelligent control method for hot dry noodle shaking equipment according to claim 2 is characterized in that: The process of obtaining the distribution uniformity includes: The stretch degree of each connected domain is calculated, and the mean of the stretch degrees of all connected domains is used as the distribution uniformity; the stretch degree satisfies the relationship: ; In the formula, For the The stretch degree of the connected domain, For the The standard deviation of the chain code increments for all chain code values ​​in a connected domain, For the The total number of pixels contained in a connected domain, For the The number of chain codes in connected domains, For the The first The chaincode increment for each chaincode.

4. The intelligent control method for hot dry noodle shaking equipment according to claim 3 is characterized in that: After obtaining the distribution uniformity, the method further includes optimizing the distribution uniformity to obtain an optimized distribution uniformity, wherein the optimized distribution uniformity satisfies the following relationship: ; In the formula, is the uniformity of the distribution after optimization, For the The number of centers whose distances from the centers of connected domains are less than the distribution radius. is the total number of connected domains, For the The total number of pixels contained in a connected domain, is the average center point density under uniform distribution, represents the natural exponential function.

5. The intelligent control method for hot dry noodle shaking equipment according to claim 4 is characterized in that: Average center point density under uniform distribution The relationship is satisfied: ; In the formula, is the total number of connected domains, is the total number of connected domains, is the number of all pixels contained in the edge image, is the distribution radius; the average value of the number of pixels contained in all connected domains is taken as the distribution radius.

6. The intelligent control method for hot dry noodle shaking equipment according to claim 5, characterized in that: The degree of jitter of the historical hot-dry noodles image satisfies the following relationship: ; In the formula, The degree of looseness of the hot dry noodles. is the uniformity of edge distribution, is the maximum value of the uniformity of edge distribution, is the minimum value of edge distribution uniformity.

7. The intelligent control method for hot dry noodle shaking equipment according to claim 6, characterized in that: The edge detection adopts the Canny edge detection algorithm.

8. An intelligent control system for hot dry noodle shaking equipment, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an intelligent control method for a hot dry noodle shaking device according to any one of claims 1 to 7 is implemented.

Images