Method and system for eliminating bubbles in fiber cloth

Abstract

The invention discloses a fiber cloth bubble eliminating method and system. The method comprises the following steps: receiving a first cloth image of the fiber cloth; performing an image preprocessing step to perform gray-scale processing on the first cloth image, and filtering a cloth texture pattern in the gray-scale image to obtain a second cloth image; performing a bubble smoothing calculation step to binarize the second cloth image to obtain a binarized image; performing feature detection on the binarized image to select a bubble pattern from the binarized image; and according to the bubble area value of the bubble pattern and a preset area threshold value, determining whether to carry out a bubble elimination step, wherein the bubble elimination step is to control the robot arm to eliminate the bubbles of the fiber cloth. The system comprises a memory and a processor for performing the method. By means of the method for eliminating the bubbles of the fiber cloth, the bubbles of the fiber cloth can be eliminated, and then the situation that the quality of the cloth is reduced due to the bubbles of the fiber cloth is avoided.

Description

Technical Field

[0001] The invention relates to a method and system for eliminating bubbles in fiber cloth. Background Art

[0002] With the development of social economy and technology, the entertainment industry also provides people with more kinds of entertainment activities. Among them, taking a yacht to sea for marine activities, such as fishing, swimming, etc., is becoming more and more popular. Therefore, the manufacturing technology of yachts is also receiving more and more attention.

[0003] In the process of manufacturing yachts, fiber cloth (e.g., glass fiber cloth) needs to be pre-impregnated with glue and then covered and adhered to the model to facilitate the manufacture of fiber cloth products for yachts. However, in the process of manufacturing fiber cloth products, air may be trapped between the fiber cloth and the model because the fiber cloth pre-impregnated with glue is covered on the model, thereby generating bubbles. The bubbles in the fiber cloth will reduce the quality of the fiber cloth products, for example, the fiber cloth is easily brittle.

[0004] Therefore, a fiber cloth bubble elimination method and system are needed to eliminate bubbles generated by the fiber cloth covering operation. Summary of the invention

[0005] The embodiments of the present invention provide a method and system for eliminating bubbles in fiber cloth, which can eliminate bubbles in fiber cloth and prevent the quality of fiber cloth products from being reduced.

[0006] According to some embodiments of the present invention, the fiber cloth bubble elimination method of the present invention comprises: receiving a first cloth image of the fiber cloth, wherein the fiber cloth is covered on the target object; performing an image pre-processing step to filter the cloth texture pattern in the first cloth image to obtain a second cloth image; performing a bubble smoothing calculation step to binarize the second cloth image to obtain a binarized image; performing a feature detection step on the binarized image to select a bubble pattern from the binarized image and calculate the bubble coordinate value and bubble area value of the bubble pattern, wherein the number of bubble coordinate values ​​and the number of bubble area values ​​correspond to the number of bubble patterns; and judging whether to perform a bubble elimination step according to the bubble area value and a preset area threshold. The above-mentioned image pre-processing step comprises: performing a grayscale processing step on the first cloth image to obtain a grayscale image; and performing a frequency domain analysis processing step and a filtering step on the grayscale image to filter the cloth texture pattern in the grayscale image to obtain the second cloth image. The above-mentioned binarized image comprises a bubble pattern, and the bubble pattern corresponds to the bubble of the fiber cloth. The above-mentioned bubble elimination step includes: calculating a robot arm moving path according to the bubble coordinate value; and controlling the robot arm to move along the robot arm moving path to eliminate the bubbles on the fiber cloth.

[0007] In some embodiments, the bubble smoothing step further includes: performing at least one dilation processing step and at least one erosion processing step on the binary image to filter out noise in the second cloth image.

[0008] In some embodiments, when the number of bubble patterns selected in the feature detection step is greater than two, the number of bubble coordinate values ​​calculated in the feature detection step is greater than two, and the number of bubble area values ​​calculated in the feature detection step is greater than two, the step of determining whether to perform a bubble elimination step based on the bubble area value and a preset area threshold includes: determining whether the bubble area value of each bubble pattern is greater than the preset area threshold; and performing a bubble elimination step when the bubble area value of one of the bubble patterns is greater than the preset area threshold.

[0009] In some embodiments, the aforementioned step of determining whether to perform the bubble elimination step based on the bubble area value and the preset area threshold further includes: marking a plurality of bubble patterns to be processed from the aforementioned bubble pattern, wherein the bubble area value of each of these bubble patterns to be processed is greater than the preset area threshold.

[0010] In some embodiments, the aforementioned step of calculating the robot arm movement path based on the bubble coordinate values ​​includes: calculating the robot arm movement path based on multiple bubble coordinate values ​​to be processed corresponding to the bubble pattern to be processed, wherein the bubble coordinate value to be processed is a plurality of the aforementioned bubble coordinate values.

[0011] According to some embodiments of the present invention, a fiber cloth bubble elimination system includes: a memory and a processor. The memory is used to store a plurality of instructions; the processor is used to load the above instructions to perform: receiving a first cloth image of a fiber cloth, wherein the fiber cloth is covered on a target object; performing an image pre-processing step to filter the cloth texture pattern in the first cloth image to obtain a second cloth image; performing a bubble smoothing calculation step to binarize the second cloth image to obtain a binarized image; performing a feature detection step on the binarized image to select a bubble pattern from the binarized image and calculate the bubble coordinate value and bubble area value of the bubble pattern, wherein the number of bubble coordinate values ​​and the number of bubble area values ​​correspond to the number of bubble patterns; and determining whether to perform a bubble elimination step according to the bubble area value and a preset area threshold. The above image pre-processing step includes: performing a grayscale processing step on the first cloth image to obtain a grayscale image; and performing a frequency domain analysis processing step and a filtering step on the grayscale image to filter the cloth texture pattern in the grayscale image to obtain the second cloth image. The binary image includes a bubble pattern, which corresponds to the bubbles on the fiber cloth. The bubble elimination step includes: calculating a robot arm movement path according to the bubble coordinate value; and controlling the robot arm to move along the robot arm movement path to eliminate the bubbles on the fiber cloth.

[0012] In some embodiments, the bubble smoothing step further comprises: performing at least one dilation processing step and at least one erosion processing step on the binary image to filter noise in the second cloth image.

[0013] In some embodiments, when the number of bubble patterns selected in the feature detection step is greater than two, the number of bubble coordinate values ​​calculated in the feature detection step is greater than two, and the number of bubble area values ​​calculated in the feature detection step is greater than two, the step of determining whether to perform a bubble elimination step based on the bubble area value and a preset area threshold includes: determining whether the bubble area value of each bubble pattern is greater than the preset area threshold; and performing a bubble elimination step when the bubble area value of one of the bubble patterns is greater than the preset area threshold.

[0014] In some embodiments, the aforementioned step of determining whether to perform the bubble elimination step based on the bubble area value and the preset area threshold further includes: marking a plurality of bubble patterns to be processed from the aforementioned bubble pattern, wherein the bubble area value of each of these bubble patterns to be processed is greater than the preset area threshold.

[0015] In some embodiments, the aforementioned step of calculating the robot arm movement path based on the bubble coordinate values ​​includes: calculating the robot arm movement path based on multiple bubble coordinate values ​​to be processed corresponding to the bubble pattern to be processed, wherein the bubble coordinate value to be processed is a plurality of the aforementioned bubble coordinate values. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a functional block diagram of a fiber cloth bubble elimination system 100 according to an embodiment of the present invention;

[0017] Figure 2 is a schematic diagram illustrating a flow chart of a fiber cloth bubble elimination method 200 according to an embodiment of the present invention;

[0018] Figure 3 is a first cloth image showing a fiber cloth according to an embodiment of the present invention;

[0019] Figure 4 is a second cloth image showing a fiber cloth according to an embodiment of the present invention;

[0020] Figure 5 is a binary image of a fiber cloth according to an embodiment of the present invention;

[0021] Figure 6 to Figure 7 1. It illustrates a dilation process step and an erosion process step according to an embodiment of the present invention;

[0022] Figure 8 FIG. 4 is a diagram illustrating the result of bubble pattern detection according to an embodiment of the present invention.

[0023]

Explanation of symbols

[0024] 100: Fiber cloth bubble elimination system

[0025] 110: Memory

[0026] 120: Processor

[0027] 200: Fiber cloth bubble elimination method

[0028] 210~250: Operation

[0029] 222-224: Operation

[0030] 232-234: Operation

[0031] 300: Fiber cloth

[0032] 300B: Binarized image

[0033] 310G: Grayscale image

[0034] 320: Second fabric image

[0035] 510: Bubble pattern

[0036] 510C: Center point

[0037] TE: Fabric texture pattern DETAILED DESCRIPTION

[0038] Please also refer to Figure 1 and Figure 2 , Figure 1 is a functional block diagram of a fiber cloth bubble elimination system 100 according to an embodiment of the present invention. Figure 2 FIG. 2 is a flow chart of a fiber cloth bubble elimination method 200 according to an embodiment of the present invention. The fiber cloth bubble elimination system 100 includes a memory 110 and a processor 120. The memory 110 is used to store a plurality of instructions for the processor 120 to load and execute. Figure 2 Fiber cloth bubble elimination method 200.

[0039] The fiber cloth bubble elimination method 200 is applicable to a fiber cloth covered on a model to detect whether there are bubbles in the fiber cloth and eliminate the bubbles in the fiber cloth. In the fiber cloth bubble elimination method 200, operation 210 is first performed to receive a first cloth image 310 of the fiber cloth 300, such as Figure 3 In this embodiment, the fiber cloth is glass fiber cloth, but the embodiments of the present invention are not limited thereto.

[0040] The fiber cloth 300 is pre-impregnated with glue and covered on a target object (e.g., a model). Since the fiber cloth 300 is covered on the target object in a wet state, air may be retained between the fiber cloth 300 and the target object, causing the fiber cloth 300 to bulge. In this document, these portions of the cloth that bulge due to the retained air are referred to as "bubbles". These bubbles can cause the fiber cloth to become brittle, so it is necessary to identify the bubbles and deal with them.

[0041] Operation 220 is an image pre-processing step to filter the texture of the fiber cloth 300 to prevent the texture of the fiber cloth 300 from affecting the bubble recognition effect. In operation 220, operation 222 is first performed to grayscale the first fabric image 310 to obtain a grayscale image 310G. Figure 3 As shown, the grayscale image 310G includes a plurality of cloth texture patterns TE. Then, operation 224 is performed to perform a frequency domain analysis processing step and a filtering step on the grayscale image to filter or fade the cloth texture patterns TE in the grayscale image 310G, thereby obtaining a second cloth image 320, as shown in FIG. Figure 4In some embodiments, the image pre-processing step uses a filter mask to filter the cloth texture pattern in the grayscale image 310G, but the embodiments of the present invention are not limited thereto.

[0042] Then, operation 230 is performed to perform a bubble smoothing calculation step. In operation 230, operation 232 is first performed to perform a binarization step on the second fabric image 320 to obtain a binarized image 300B. Figure 5 As shown. In some embodiments, before the binarization step is performed, image blurring may also be performed. The binary image 300B includes at least one bubble pattern 510, and the bubble pattern 510 corresponds to the bubbles of the fabric fiber cloth 300. In other words, the bubble pattern 510 is the pattern of bubbles of the fiber cloth 300 displayed in the image. Then, operation 234 is performed to perform a dilation processing step and an erosion processing step on the binary image to remove noise in the image. In this embodiment, the binary image 300B is subjected to two dilation processing steps and two erosion processing steps, as shown in FIG. Figure 6 to Figure 7 As shown. This can effectively remove the noise in the binary image 300B and facilitate the detection of the bubble pattern 510 in the binary image 300B. Hereinafter, the binary image from which the noise has been removed is referred to as the third fabric image.

[0043] Next, operation 240 is performed to perform a feature detection step on the third fabric image. The feature detection step can detect (select) a bubble pattern from the third fabric image, such as Figure 8 Operation 240 is to detect the bubble pattern in the third fabric image using an object detection algorithm, such as a convolutional neural network (CNN), a region convolutional neural network (RCNN), YOLO, etc., to select the bubble pattern 510 and calculate the coordinates and area size of each bubble pattern 510.

[0044] Specifically, if Figure 8 As shown, operation 240 is to detect multiple bubble patterns 510, and frame and mark the bubble patterns 510. At the same time, the center point 510C of each bubble pattern 510 is calculated, and the coordinate value of the center point 510C is used to represent the coordinate value of the bubble pattern 510. In addition, in some embodiments, the area size of each bubble pattern 510 is simultaneously displayed in the upper left corner of the framed bubble pattern 510, for example, size=3478.

[0045] In some embodiments, operation 234 may be omitted. In other words, the binary image 300B does not necessarily need to undergo dilation and erosion. Operation 240 may directly perform feature extraction on the binary image 300B.

[0046] Then, operation 250 is performed to determine whether to perform the bubble elimination step according to the bubble area value and the preset area threshold. For example, in some embodiments, the preset area threshold is a threshold corresponding to a single bubble. When the area of ​​a bubble pattern 510 is greater than the preset area threshold, it means that the bubble corresponding to this bubble pattern 510 is too large and needs to be processed. Therefore, operation 250 compares the area value of each bubble pattern 510 with the preset area threshold to find the bubble pattern 510 to be processed and record its coordinate value.

[0047] The bubble elimination step of the embodiment of the present invention is performed by using a robot arm. For example, in the bubble elimination step of the embodiment, the movement path of the robot arm is first calculated according to the coordinate values ​​of all the bubble patterns 510 to be processed, and then the robot arm is moved according to the movement path to roll on the coordinate values ​​of the bubble patterns 510 to be processed to eliminate bubbles.

[0048] In some embodiments of the present invention, operation 250 may determine whether to perform a bubble elimination step according to the regional bubble density on the fiber cloth 300. For example, it is calculated whether the sum of the bubble areas in the preset area is greater than a preset area threshold. If the sum of the bubble areas in the preset area is greater than the preset area threshold, it means that the bubble density in the preset area is large, so the bubble elimination step is performed to eliminate the bubbles in the preset area.

[0049] From the above description, it can be seen that the embodiment of the present invention provides a fiber cloth bubble elimination method and system, which uses a custom binary mask smoothing algorithm to analyze the subtle bubble features on the fiber cloth. After identifying and dialysing the bubble area on the fiber cloth, the bubble position coordinates and bubble area size values ​​are generated on the bubble analysis area, and the bubble feature .csv value file (i.e. the movement path of the above-mentioned robot arm) is generated, which can be used by the robot arm to perform the bubble elimination operation.

[0050] The fiber cloth bubble elimination method and system of the embodiment of the present invention can be used in fiber cloth products of small working ships, and can also be matched with the hull of medium-sized yachts to reduce the forming defects of fiber cloth products and reduce the repeated manual rework and repair process. In addition to the small and medium-sized ship industry, it can also be applied to products such as surfboards, paddle boards and high-end helmets in the sports industry.

[0051] Although the present invention has been disclosed as above by way of embodiments, it is not intended to limit the present invention. Any person having ordinary knowledge in the technical field may make some changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be determined by the definition of the attached claims.

Claims

1. A fiber cloth bubble elimination method, applicable to a fiber cloth bubble elimination system, It is characterized in that Include: Receiving a first cloth image of a fiber cloth, wherein the fiber cloth covers a target object; Perform an image pre-processing step, including: Performing a grayscale processing step on the first cloth image to obtain a grayscale image, wherein the grayscale image includes a plurality of cloth texture patterns; and Performing a frequency domain analysis processing step and a filtering step on the grayscale image to filter the plurality of cloth texture patterns in the grayscale image to obtain a second cloth image; A bubble smoothing step is performed to obtain: Performing a binarization step on the second fabric image to obtain a binarized image, wherein the binarized image includes at least one bubble pattern, and the at least one bubble pattern corresponds to at least one bubble of the fiber cloth; Performing a feature detection step on the binary image to select the at least one bubble pattern from the binary image, and calculating at least one bubble coordinate value and at least one bubble area value of the at least one bubble pattern, wherein the number of the at least one bubble coordinate value and the number of the at least one bubble area value correspond to the number of the at least one bubble pattern; and Determining whether to perform a bubble elimination step according to the at least one bubble area value and a preset area threshold, wherein the bubble elimination step includes: Calculating a robot arm movement path according to the at least one bubble coordinate value; as well as The robot arm is controlled to move along the robot arm movement path to eliminate the at least one bubble of the fiber cloth.

2. The method for eliminating bubbles from a fiber cloth as claimed in claim 1, It is characterized in that The bubble smoothing calculation step also includes: The binary image is subjected to at least one dilation processing step and at least one erosion processing step to filter noise in the second cloth image.

3. The method for eliminating bubbles from a fiber cloth as claimed in claim 1, It is characterized in that When the number of the at least one bubble pattern selected by the feature detection step is two or more, the number of the at least one bubble coordinate value calculated by the feature detection step is two or more, and the number of the at least one bubble area value calculated by the feature detection step is two or more, the step of determining whether to perform the bubble elimination step according to the at least one bubble area value and the preset area threshold comprises: Determine whether the area value of each of the plurality of bubbles in each of the plurality of bubble patterns is greater than the preset area threshold; When the bubble area value of one of the plurality of bubble patterns is greater than the preset area threshold, the bubble elimination step is performed.

4. The method for eliminating bubbles from a fiber cloth as claimed in claim 3, It is characterized in that The step of determining whether to perform the bubble elimination step according to the at least one bubble area value and the preset area threshold further comprises: A plurality of bubble patterns to be processed are marked from the plurality of bubble patterns, wherein the bubble area value of each of the plurality of bubble patterns to be processed is greater than the preset area threshold.

5. The method for eliminating bubbles from a fiber cloth as claimed in claim 4, It is characterized in that The step of calculating the movement path of the robot arm according to the at least one bubble coordinate value comprises: The movement path of the robot arm is calculated according to a plurality of bubble coordinate values ​​to be processed corresponding to the plurality of bubble patterns to be processed, wherein the plurality of bubble coordinate values ​​to be processed are a plurality of the plurality of bubble coordinate values.

6. A fiber cloth bubble elimination system, It is characterized in that Include: a memory for storing a plurality of instructions; A processor is configured to load the plurality of instructions to perform: Receiving a first cloth image of a fiber cloth, wherein the fiber cloth covers a target object; Perform an image pre-processing step, including: Performing a grayscale processing step on the first cloth image to obtain a grayscale image, wherein the grayscale image includes a plurality of cloth texture patterns; and Performing a frequency domain analysis processing step and a filtering step on the grayscale image to filter the plurality of cloth texture patterns in the grayscale image to obtain a second cloth image; A bubble smoothing calculation step is performed, including: Performing a binarization step on the second fabric image to obtain a binarized image, wherein the binarized image includes at least one bubble pattern, and the at least one bubble pattern corresponds to at least one bubble of the fiber cloth; as well as Performing a feature detection step on the binary image to select the at least one bubble pattern from the binary image, and calculating at least one bubble coordinate value and at least one bubble area value of the at least one bubble pattern, wherein the number of the at least one bubble coordinate value and the number of the at least one bubble area value correspond to the number of the at least one bubble pattern; Determining whether to perform a bubble elimination step according to the at least one bubble area value and a preset area threshold, wherein the bubble elimination step includes: Calculating a robot arm movement path according to the at least one bubble coordinate value; as well as The robot arm is controlled to move along the robot arm movement path to eliminate the at least one bubble of the fiber cloth.

7. The fiber cloth bubble elimination system according to claim 6, It is characterized in that The bubble smoothing calculation step also includes: The binary image is subjected to at least one dilation processing step and at least one erosion processing step to filter noise in the binary image.

8. The fiber cloth bubble elimination system according to claim 6, It is characterized in that When the number of the at least one bubble pattern selected by the feature detection step is two or more, the number of the at least one bubble coordinate value calculated by the feature detection step is two or more, and the number of the at least one bubble area value calculated by the feature detection step is two or more, the step of determining whether to perform the bubble elimination step according to the at least one bubble area value and the preset area threshold comprises: Determine whether the area value of each of the plurality of bubbles in each of the plurality of bubble patterns is greater than the preset area threshold; When the bubble area value of one of the plurality of bubble patterns is greater than the preset area threshold, the bubble elimination step is performed.

9. The fiber cloth bubble elimination system as claimed in claim 8, It is characterized in that The step of determining whether to perform the bubble elimination step according to the at least one bubble area value and the preset area threshold further comprises: A plurality of bubble patterns to be processed are marked from the plurality of bubble patterns, wherein the bubble area value of each of the plurality of bubble patterns to be processed is greater than the preset area threshold.

10. The fiber cloth bubble elimination system according to claim 9, It is characterized in that The step of calculating the movement path of the robot arm according to the at least one bubble coordinate value comprises: The movement path of the robot arm is calculated according to a plurality of bubble coordinate values ​​to be processed corresponding to the plurality of bubble patterns to be processed, wherein the plurality of bubble coordinate values ​​to be processed are a plurality of the plurality of bubble coordinate values.

Images