Ultrasonic cutting detection and analysis system based on visual data analysis

Abstract

The present invention discloses an ultrasonic cutting detection and analysis system based on visual data analysis, which belongs to the field of ultrasonic cutting detection technology and comprises a processing analysis module, an acquisition module and a detection module; the processing analysis module is used for performing processing analysis, acquiring target information, and determining the cutting position on the processing material in real time according to the target information; and performing line light marking on the cutting position through a light line unit; the acquisition module is used for performing real-time image acquisition during the ultrasonic cutting process to obtain a corresponding cutting detection image; the detection module is used for detecting the cutting detection image, identifying the position of the cutter head in the cutting detection image in real time, and marking it as a base point position; generating a detection area with the base point position as the center, performing grayscale processing on the cutting detection image within the detection area, and obtaining a detection grayscale image; establishing a detection model according to a material model and the detection grayscale image; performing real-time detection evaluation according to the detection model, and obtaining a corresponding detection evaluation result.

Description

Technical Field

[0001] The present invention belongs to the technical field of ultrasonic cutting detection, and in particular to an ultrasonic cutting detection and analysis system based on visual data analysis. Background Art

[0002] In modern industrial production, cutting technology, as one of the key manufacturing processes, is crucial to product molding, quality and production efficiency. However, in the cutting process, the existing cutting methods are prone to adhesion, burrs and other problems, resulting in the cutting effect not reaching the ideal state; with the development of technology, ultrasonic cutting has emerged. Ultrasonic cutting does not require a sharp blade, reduces material damage and edge collapse, and has a fusion effect at the same time, which can prevent the loosening of the cut material structure and has outstanding application effects.

[0003] However, ultrasonic cutting also has certain shortcomings. For example, in some materials, especially those containing impurities, uneven distribution or complex internal structures, problems such as uneven cuts and excessive cuts are prone to occur during the cutting process. In addition, environmental factors in the cutting process are also factors that cannot be ignored. For example, changes in environmental factors such as temperature, humidity, and vibration may affect the cutting results and thus affect the cutting quality.

[0004] Based on this, in order to improve the cutting quality, the present invention provides an ultrasonic cutting detection and analysis system based on visual data analysis. Summary of the invention

[0005] In order to solve the problems existing in the above-mentioned solutions, the present invention provides an ultrasonic cutting detection and analysis system based on visual data analysis.

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

[0007] Ultrasonic cutting detection and analysis system based on visual data analysis, including processing analysis module, acquisition module and detection module;

[0008] The processing analysis module is used to perform processing analysis, obtain target information corresponding to the processing target, determine the cutting position on the processing material in real time according to the target information; and perform line light marking on the cutting position through the light line unit.

[0009] Furthermore, the method for determining the cutting position includes:

[0010] subjecting the processing material to initial conditioning;

[0011] Establishing a material model and a target model corresponding to the processing material and the processing target, respectively, superimposing the target model into the material model, and making corresponding position changes to the target model in the material model, estimating the corresponding material waste in real time, marking the target model position corresponding to the minimum material waste as the optimal position, and adjusting the target model to the optimal position; forming a combined model consisting of the material model and the target model;

[0012] Determining a cutting boundary corresponding to the processing target according to the combined model;

[0013] The image of the processed material is acquired in real time, and the cutting boundary is located in the processed material according to the image of the processed material to determine the cutting position.

[0014] Furthermore, the method for locating the cutting boundary according to the processed material image includes:

[0015] Perform grayscale processing on the processed material image to obtain the material grayscale image, and identify the grayscale value of each pixel in the material grayscale image;

[0016] Calculate the difference value of each pixel according to the gray value of each pixel; calculate the absolute value of the difference between the difference values ​​of each adjacent pixel, and mark it as the combined analysis value;

[0017] The corresponding positions of two adjacent pixels whose combined analysis values ​​are less than the threshold value X1 are combined to obtain a combined area; the average value of the difference values ​​corresponding to the pixels in the combined area is taken as the difference value of the combined area;

[0018] Calculate the merged analysis value of pixels adjacent to the merged area or between merged areas; merge the corresponding positions of adjacent pixels or merged areas whose merged analysis values ​​are less than the threshold value X1 to obtain a new merged area; and so on until merging is no longer possible; mark the corresponding positions of the remaining pixels and the corresponding positions of the merged area as initial areas;

[0019] Identify the area and difference value of each initial area; screen each initial area according to the obtained difference value and area to obtain each positioning area;

[0020] The control area of ​​each positioning area in the combined model is identified, and the cutting position is determined according to each positioning area and the control area.

[0021] Furthermore, the calculation method of the difference value includes:

[0022] Calculate the gray value difference between each pixel and its adjacent pixels, and mark the gray value difference as di(x, y), where (x, y) represents the position of the corresponding pixel; i represents the adjacent pixel of the corresponding pixel, i=1, 2, ..., n, and n is a positive integer;

[0023] According to the formula Calculate the difference value of each pixel;

[0024] Where: CP(x, y) is the difference value.

[0025] Furthermore, the working method of the lamp line unit includes:

[0026] Identify each candidate color, illuminate each candidate color one by one on the processing material according to the preset path, and perform corresponding image acquisition to obtain the path acquisition image corresponding to each candidate color; compare the clarity of each path acquisition image, and mark the candidate color corresponding to the path acquisition image with the highest clarity as the target color;

[0027] Acquire processing equipment information, and set a corresponding reference width and a seam width standard range according to the processing equipment information; determine the width of the lamp line according to the seam width standard range;

[0028] Adjust the light parameters according to the target color and light line width obtained;

[0029] Identify the cutting position and mark the cutting position with line light according to the set light parameters.

[0030] Furthermore, there is no target color on the processed material.

[0031] The acquisition module is used to perform real-time image acquisition during the ultrasonic cutting process to obtain corresponding cutting detection images.

[0032] The detection module is used to detect the cutting detection image, identify the position of the tool head in the cutting detection image in real time, and mark it as the base point position; generate a detection area based on the base point position as the center, perform grayscale processing on the cutting detection image in the detection area, and obtain a detection grayscale image; identify the grayscale value of each pixel in the detection grayscale image;

[0033] Acquire the material model, analyze the detection grayscale image, and obtain the line light position, line light width, slit position, and slit width at the corresponding time; mark the obtained line light position, line light width, slit position, and slit width in the material model in real time, and mark the current material model as the detection model;

[0034] Perform real-time detection and evaluation based on the detection model to obtain corresponding detection and evaluation results.

[0035] Furthermore, the method for analyzing the detected grayscale image includes:

[0036] Establish a line light recognition model, the expression of the line light recognition model is ; In the formula: q is the input data, the input data is the gray value of the pixel; the output data is the line light judgment value 1 or 0;

[0037] The grayscale value of each pixel in the detection grayscale image is analyzed by the line light recognition model to obtain the line light judgment value corresponding to each pixel, and the line light position is marked in the detection grayscale image according to the line light judgment value to identify the line light width;

[0038] Establish a cutting judgment model. The expression of the cutting judgment model is: ; Where: s is the input data, which includes the grayscale value of the pixel to be evaluated and the grayscale values ​​of each pixel adjacent to the pixel; the output data is the cropping judgment value 2 or 0;

[0039] The detection grayscale image is analyzed through the cropping judgment model to obtain the cropping judgment value of each pixel. The cutting position is marked in the detection grayscale image according to the cropping judgment value to identify the cutting width.

[0040] Furthermore, the method for performing real-time detection evaluation based on the detection model includes:

[0041] Calculate the deviation between the slit width and the line light width at the corresponding position in real time; identify the detection time corresponding to the deviation value;

[0042] Generate a corresponding detection curve according to the detection time corresponding to each deviation value, where the horizontal axis of the detection curve is the detection time and the vertical axis is the deviation value; fit the detection curve to obtain a corresponding detection function; mark the detection function as KH(t), where t is the detection time;

[0043] According to the formula Calculate the corresponding detection evaluation value;

[0044] Where: PQ is the detection evaluation value;

[0045] An evaluation is performed based on the obtained detection evaluation value to obtain a corresponding detection evaluation result.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] Intelligent detection and analysis of ultrasonic cutting is achieved through the cooperation between the processing analysis module, the acquisition module and the detection module; and based on the light line unit, the detection area, etc., the data processing volume is reduced and the analysis speed is improved; the cutting data can be analyzed in real time, and abnormal situations such as incision deviation and excessive cutting seam can be discovered in time, and alarms can be issued to operators to reduce losses; it can effectively reduce the production of defective products and improve product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 This is a principle block diagram of the present invention. DETAILED DESCRIPTION

[0050] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] like Figure 1 As shown, the ultrasonic cutting detection and analysis system based on visual data analysis includes a processing analysis module, a collection module and a detection module;

[0052] The processing analysis module is used to perform processing analysis and obtain target information corresponding to the processing target. The processing target refers to the product that needs to be processed and cut. The target information includes information such as shape and size; the cutting position on the processing material is determined in real time according to the target information. The processing material refers to the whole material before processing, such as the whole fabric, and the finished product is the product that needs to be cut from the whole fabric; and the light line unit is used to perform line light marking on the cutting position.

[0053] The method for determining the cutting position on the processing material in real time according to the target information includes:

[0054] Make initial adjustments to the processed materials. Initial adjustments refer to adjustments made after each cutting of a product. For example, after each cutting, the material is dragged to a preset length for subsequent cutting. The specific adjustments are made according to the working mode of the ultrasonic cutting equipment.

[0055] Establish material models and target models corresponding to processing materials and processing targets respectively. The material model and target model can be two-dimensional or three-dimensional display models. The material model only needs to include a part of the material that is more than the processing target, and does not need the entire processing material. It can meet the display of subsequent processing targets and cutting positions. The target model is established according to the target information; the target model is superimposed on the material model, and the corresponding position optimization adjustment is performed to obtain a combined model. The position optimization adjustment is adjusted based on the minimum material waste. The corresponding material waste can be directly identified based on the existing technology, and its optimal position is determined according to the minimum material waste, and the target model is adjusted to the corresponding position;

[0056] Determine the corresponding cutting boundary for cutting the processing target according to the combination model;

[0057] Acquire the image of the processed material in real time, mainly the image of the corresponding part of the combined model; locate the cutting boundary according to the image of the processed material, determine the position of the cutting boundary on the processed material in real time, and mark it as the cutting position. It is also not necessary to determine the entire cutting boundary, as long as the position of the front section of the cutter head is guaranteed, and then determine it later.

[0058] Methods for locating the cutting boundary based on the processed material image include:

[0059] Perform grayscale processing on the processed material image to obtain the material grayscale image, and identify the grayscale value of each pixel in the material grayscale image;

[0060] Calculate the gray value difference between each pixel and its adjacent pixels, and mark the gray value difference as di(x, y), where (x, y) represents the position of the corresponding pixel, that is, which pixel is used as the reference for calculation with the adjacent pixels; i represents the adjacent pixel of the corresponding pixel, i=1, 2, ..., n, and n is a positive integer;

[0061] According to the formula Calculate the difference value of each pixel; where: is the difference value;

[0062] Calculate the absolute value of the difference between the difference values ​​of each adjacent pixel and mark it as the combined analysis value;

[0063] The corresponding positions of two adjacent pixels whose combined analysis values ​​are less than the threshold value X1 are combined to obtain a combined area; the average value of the difference values ​​corresponding to the pixels in the combined area is taken as the difference value of the combined area;

[0064] Calculate the merged analysis value of pixels adjacent to the merged area or between merged areas; merge the corresponding positions of adjacent pixels or merged areas whose merged analysis values ​​are less than the threshold value X1 to obtain a new merged area; and so on until merging is no longer possible; mark the corresponding positions of the remaining pixels and the corresponding positions of the merged area as initial areas;

[0065] Identify the area and difference value of each initial area; screen each initial area according to the obtained difference value and area, and mark the remaining initial areas as positioning areas; that is, eliminate the initial areas whose difference value is lower than the preset value and whose area is not within the preset range, and set the corresponding preset value according to whether the angle is easy to identify and locate; in actual applications, a large number of images of processed materials will be collected for simulated positioning, and then the corresponding preset value for screening will be determined.

[0066] Identify the corresponding position of each positioning area in the combined model, mark it as the reference area, determine the position of the cutting boundary in the processed image based on each positioning area and the reference area, and then determine the corresponding position on the processed material, which is marked as the cutting position. In practical applications, due to wrinkles in the processed material, multiple positioning areas can eliminate adverse effects; achieve accurate positioning of the cutting boundary, and then determine the cutting position corresponding to the cutting boundary on the current processed material.

[0067] The light line unit is an existing device that can emit light at a specified position according to needs, and the light width and light color are adjustable, such as a red, thin light. The light line unit can adjust the light parameters manually or in an intelligent way. The light parameters refer to parameters such as light width, color, and brightness. The method of intelligent adjustment includes:

[0068] Identify the various light colors of the light line unit, mark them as candidate colors, illuminate the candidate colors one by one on the processing material according to the preset path, and perform corresponding image acquisition to obtain the path acquisition image corresponding to each candidate color; compare and evaluate the clarity of the paths in the path acquisition diagram to determine the path acquisition image with the highest effect. A good effect means that the grayscale values ​​of its imaging pixels have a relatively obvious difference, and the grayscale value can be calculated to determine whether it belongs to the line light color; mark the corresponding candidate color as the target color; and generally eliminate the candidate colors corresponding to the colors in the processing material to facilitate the identification of line light;

[0069] Obtain processing equipment information, that is, relevant information of ultrasonic cutting equipment, mainly the cutting width under normal conditions, that is, the seam width formed after cutting by the cutter head, marked as the reference width; also includes the allowable width error; then the seam width range can be determined according to the width error and the reference width in the processing equipment information, marked as the seam width standard range; the minimum seam width in the seam width standard range is the light line width;

[0070] Adjust the light parameters according to the target color and light line width.

[0071] The acquisition module is used to perform real-time image acquisition of the cutting path during the ultrasonic cutting process to obtain a corresponding cutting detection image.

[0072] The detection module is used to detect the cutting detection image, identify the position of the cutter head in the cutting detection image in real time, and mark it as the base point position; generate a detection area based on the base point position as the center, that is, including the base point position, the cutting seam and other areas, and preset a radius for direct generation to reduce the amount of data processing; grayscale processing is performed on the cutting detection image in the detection area to obtain a detection grayscale image; identify the grayscale value of each pixel in the detection grayscale image;

[0073] A line light recognition model is established. The line light recognition model is used to intuitively identify the difference between the target color corresponding to the line light and the pixels corresponding to the processed material; the pixels belonging to the target color are regarded as line light pixels; the expression of the line light recognition model is ; In the formula: q is the input data, the input data is the gray value of the pixel; the output data is the line light judgment value; because when selecting the target color of the line light, the gray value selected is significantly different from the color of the processed material, so the judgment can be made based on the gray difference;

[0074] Obtain the material model, analyze the grayscale value of each pixel in the detection grayscale image through the line light recognition model, obtain the line light judgment value corresponding to each pixel, mark the line light position in the detection grayscale image according to the line light judgment value, and identify the line light width; mark the line light in the material model according to the obtained line light position and line light width;

[0075] A cutting judgment model is established. The cutting judgment model is used to judge whether the pixel is within the cutting width. A gap is formed after cutting based on the grayscale difference and the grayscale difference of the light line. The grayscale difference of the light line is the grayscale value difference between the target color of the line light and the corresponding color of the material. It can be judged whether the pixel is within the cutting width. If it is within the cutting width, it is considered to meet the cutting judgment criteria. The expression of the cutting judgment model is: ; Where: s is the input data, which includes the grayscale value of the pixel to be evaluated and the grayscale values ​​of each pixel adjacent to the pixel; the output data is the cropping judgment value 2 or 0;

[0076] The detection grayscale image is analyzed by the cutting judgment model to obtain the cutting judgment value of each pixel, and the cutting position is marked in the detection grayscale image according to the cutting judgment value to identify the cutting width; the cutting is marked in the material model according to the cutting position and the cutting width;

[0077] Mark the current material model as a detection model; perform real-time detection and evaluation based on the detection model to obtain corresponding detection and evaluation results.

[0078] Methods for real-time detection evaluation based on detection models include:

[0079] Calculate the deviation between the slit width and the line light width at the corresponding position in real time, because the cutter head should cut along the center line of the line light. When it deviates from the center line or the slit width becomes larger, deviation will occur. Exceeding the line light range is regarded as deviation. The size exceeding the line light range is the deviation value. If one side exceeds, it means that one side exceeds the size, and one side is specified as positive and the other side is negative; if both sides exceed, it is the sum of the sizes that exceed on both sides, and the sign of the side with more excess shall prevail. If the excess on both sides is equal, the sign is the same as the previous time; identify the detection time corresponding to the deviation value, which refers to the time of shearing;

[0080] Generate a corresponding detection curve according to the detection time corresponding to each deviation value, where the horizontal axis of the detection curve is the detection time and the vertical axis is the deviation value; fit the detection curve to obtain a corresponding detection function; mark the detection function as KH(t), where t is the detection time;

[0081] According to the formula Calculate the corresponding detection evaluation value; where: PQ is the detection evaluation value; the threshold value X2 is set according to the standard range of the seam width; the threshold value X3 is set according to the user's requirements for the cumulative deviation;

[0082] According to the obtained detection evaluation value, an evaluation is performed to obtain the corresponding detection evaluation result. That is, if the detection evaluation value is 1, it means that the width exceeds the range; if the detection evaluation value is 2, it means that the cutting is unstable and exceeds the allowable range; if the detection evaluation value is 3, it means normal.

[0083] The above formulas are all calculated by removing dimensions and taking numerical values. The formula is a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained by simulating a large amount of data.

[0084] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. Ultrasonic cutting detection and analysis system based on visual data analysis, It is characterized in that It includes processing and analysis module, acquisition module and detection module; The processing analysis module is used to perform processing analysis, obtain target information corresponding to the processing target, and determine the cutting position on the processing material in real time according to the target information; and perform line light marking on the cutting position through the light line unit; The acquisition module is used to perform real-time image acquisition during the ultrasonic cutting process to obtain the corresponding cutting detection image; The detection module is used to detect the cutting detection image, identify the position of the tool head in the cutting detection image in real time, and mark it as the base point position; generate a detection area based on the base point position as the center, perform grayscale processing on the cutting detection image in the detection area, and obtain a detection grayscale image; identify the grayscale value of each pixel in the detection grayscale image; Obtain the material model, analyze the detected grayscale image, and obtain the line light position, line light width, slit position and slit width at the corresponding time; Mark the obtained line light position, line light width, slit position and slit width in the material model in real time, and mark the current material model as a detection model; Perform real-time detection and evaluation according to the detection model to obtain corresponding detection and evaluation results; Methods for analyzing the detection grayscale image include: Establish a line light recognition model, the expression of the line light recognition model is ; In the formula: q is the input data, the input data is the gray value of the pixel; the output data is the line light judgment value 1 or 0; The grayscale value of each pixel in the detection grayscale image is analyzed by the line light recognition model to obtain the line light judgment value corresponding to each pixel, and the line light position is marked in the detection grayscale image according to the line light judgment value to identify the line light width; Establish a cutting judgment model. The expression of the cutting judgment model is: ; Wherein: s is the input data, which includes the grayscale value of the pixel to be evaluated and the grayscale values ​​of each pixel adjacent to the pixel to be evaluated; the output data is the cropping judgment value 2 or 0; The detection grayscale image is analyzed by the cutting judgment model to obtain the cutting judgment value of each pixel, and the cutting position is marked in the detection grayscale image according to the cutting judgment value to identify the cutting width; Methods for real-time detection evaluation based on detection models include: Calculate the deviation between the slit width and the line light width at the corresponding position in real time; identify the detection time corresponding to the deviation value; Generate a corresponding detection curve according to the detection time corresponding to each deviation value, where the horizontal axis of the detection curve is the detection time and the vertical axis is the deviation value; fit the detection curve to obtain a corresponding detection function; mark the detection function as KH(t), where t is the detection time; According to the formula Calculate the corresponding detection evaluation value; Where: PQ is the detection evaluation value; An evaluation is performed based on the obtained detection evaluation value to obtain a corresponding detection evaluation result.

2. The ultrasonic cutting detection and analysis system based on visual data analysis according to claim 1, It is characterized in that Methods for determining the cutting position include: subjecting the processing material to initial conditioning; Establishing a material model and a target model corresponding to the processing material and the processing target, respectively, superimposing the target model into the material model, and making corresponding position changes to the target model in the material model, estimating the corresponding material waste in real time, marking the target model position corresponding to the minimum material waste as the optimal position, and adjusting the target model to the optimal position; forming a combined model consisting of the material model and the target model; Determining a cutting boundary corresponding to the processing target according to the combined model; The image of the processed material is acquired in real time, and the cutting boundary is located in the processed material according to the image of the processed material to determine the cutting position.

3. The ultrasonic cutting detection and analysis system based on visual data analysis according to claim 2, It is characterized in that Methods for positioning the cutting boundary based on the processed material image include: Perform grayscale processing on the processed material image to obtain the material grayscale image, and identify the grayscale value of each pixel in the material grayscale image; Calculate the difference value of each pixel according to the gray value of each pixel; calculate the absolute value of the difference between the difference values ​​of each adjacent pixel, and mark it as the combined analysis value; The corresponding positions of two adjacent pixels whose combined analysis values ​​are less than the threshold value X1 are combined to obtain a combined area; the average value of the difference values ​​corresponding to the pixels in the combined area is taken as the difference value of the combined area; Calculate the merged analysis value of pixels adjacent to the merged area or between merged areas; merge the corresponding positions of adjacent pixels or merged areas whose merged analysis values ​​are less than the threshold value X1 to obtain a new merged area; and so on until merging is no longer possible; mark the corresponding positions of the remaining pixels and the corresponding positions of the merged area as initial areas; Identify the area and difference value of each initial area; screen each initial area according to the obtained difference value and area to obtain each positioning area; The control area of ​​each positioning area in the combined model is identified, and the cutting position is determined according to each positioning area and the control area.

4. The ultrasonic cutting detection and analysis system based on visual data analysis according to claim 3, It is characterized in that The difference value is calculated by: Calculate the gray value difference between each pixel and its adjacent pixels, and mark the gray value difference as di(x, y), where (x, y) represents the position of the corresponding pixel; i represents the adjacent pixel of the corresponding pixel, i=1, 2, ..., n, and n is a positive integer; According to the formula Calculate the difference value of each pixel; Where: CP(x, y) is the difference value.

5. The ultrasonic cutting detection and analysis system based on visual data analysis according to claim 1, It is characterized in that The working methods of the lamp line unit include: Identify each candidate color, illuminate each candidate color one by one on the processing material according to the preset path, and perform corresponding image acquisition to obtain the path acquisition image corresponding to each candidate color; compare the clarity of each path acquisition image, and mark the candidate color corresponding to the path acquisition image with the highest clarity as the target color; Acquire processing equipment information, and set a corresponding reference width and a seam width standard range according to the processing equipment information; determine the width of the lamp line according to the seam width standard range; Adjust the light parameters according to the target color and light line width obtained; Identify the cutting position and mark the cutting position with line light according to the set light parameters.

6. The ultrasonic cutting detection and analysis system based on visual data analysis according to claim 1, It is characterized in that There is no target color on the processed material.

Images