Textile fabric product defect detection method and system

Abstract

The invention discloses a textile product defect detection method and system. The method comprises the following steps: S101, shooting an intact fabric reference picture; the method comprises the following steps: smoothing and paving intact fabric products and fabric products with different defects, shooting the fabric products in multiple directions through an industrial camera and the like, and taking a shot picture as a reference contrast during detection; s102, moving the fabric product, and aligning and positioning the fabric product; the fabric product is laid on the conveyor, the conveyor is used for moving and transporting the fabric product, the fabric product is positioned, and the fabric product is detected after being positioned. According to the textile product defect detection method and system provided by the invention, the textile product is photographed and is compared with a reference and comparison textile picture for detection, and the defect type of the textile product is matched and confirmed and is detected for multiple rounds; and misjudgment caused by influence of shadow generated by indentations or wrinkles on fabric products on detection and comparison is prevented.

Description

Technical Field

[0001] The present invention relates to the technical field of textile production, and in particular to a method and system for detecting defects in textile products. Background Art

[0002] The original meaning of textile is a general term for spinning and weaving. However, with the continuous development and improvement of the textile knowledge system and discipline system, especially the emergence of technologies such as non-woven textile materials and three-dimensional composite weaving, it has expanded beyond traditional hand-spinning and weaving to include non-woven fabric technology, modern three-dimensional weaving technology, modern electrostatic nano-netting technology, and other production methods for clothing, industrial use, and decorative textiles. Therefore, modern textile refers to a multi-scale structural processing technology for fibers or fiber aggregates. During textile weaving, the warp and weft threads of the fabric are prone to breakage or holes due to uneven tension or weak thread strength during production. After the fabric product is produced, it is necessary to inspect the surface defects of the fabric product.

[0003] However, in the prior art, when inspecting fabrics, industrial cameras are mostly used for shooting and comparison. However, only one round of comparison is performed during the comparison. When the fabric is placed and moved for inspection, the fabric is prone to wrinkles due to movement or bulges and depressions due to pressing and pulling. Wrinkles and bulges and depressions are prone to shadows. Shadows are easily misjudged during inspection and comparison, and the machine may regard them as having defects, thereby classifying them as defective fabric products. Moreover, the defect category of the fabric is not judged during comparison, and there is a lack of multiple rounds of inspection of the fabric. Summary of the Invention

[0004] The object of the present invention is to provide a method and system for detecting defects in textile products, which has the functions of taking photos of textile products and comparing them with reference fabric photos, matching and confirming the types of defects in textile products and performing multiple rounds of detection, preventing shadows caused by indentations or wrinkles on textile products from affecting detection and comparison and causing misjudgment, so as to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for detecting defects in textile products, comprising the following steps: S101. Take reference pictures of intact fabrics; The intact fabric products and fabric products with different defects are smoothed and laid out, and the fabric products are photographed from multiple directions using an industrial camera, etc. The photographs are used as reference for comparison during inspection; S102, moving the fabric product and aligning and positioning it; Lay the fabric product on a conveyor, move the fabric product through the conveyor, position the fabric product, and then inspect the fabric product after positioning; S103, comparing the fabric product with a reference picture; After the fabric product is positioned, the industrial camera takes a photo of the fabric product and compares the photo with a previously taken reference photo of an intact fabric product to detect whether the fabric product has any defects. S104, analyzing and comparing whether the fabric has defects and the types of defects; When comparing fabric products, first compare them with photos of intact fabric products to detect whether there are any defects on the fabric products. If a difference is detected between the fabric products and the intact fabric products, then compare them with photos of the fabric products with defects to determine the type of defects on the fabric products. S105, performing mobile sorting on the fabrics.

[0006] After the fabric products are inspected, they are transported further via conveyors. During transportation, the intact and defective fabric products are moved and sorted by conveyors in different directions for classification and differentiation.

[0007] Furthermore, the step S101 further includes the following steps: S201, industrial camera takes pictures; S202, uploading the captured photos to the cloud; S203, setting the captured photo as a reference; Intact fabric products are laid out flat, and industrial cameras are used to photograph the intact fabric products from multiple directions and angles. After the photographing is completed, fabric products with different types of defects are laid out flat, and the defective fabric products are photographed from multiple directions and angles using industrial cameras. After the photographing is completed, the photographs are uploaded to the cloud and stored in the cloud as reference images. The stored photos are used for reference and comparison during inspection.

[0008] Furthermore, the step S102 further includes the following steps: S301, placing the fabric on a conveyor; S302, conveying the fabric through a conveyor; S303, positioning through an infrared positioning sensor; The fabric product is laid flat on the conveyor, and the conveyor drives the fabric to move and transport. When the fabric product moves, the infrared positioning sensor is used to detect the area directly below. When the fabric product moves under the infrared positioning sensor, the infrared positioning sensor locates the fabric product. The infrared positioning sensor sends a signal and controls the conveyor to pause through the control circuit, so that the fabric product stays under the infrared positioning sensor and is aligned with the industrial camera.

[0009] Furthermore, the S103 further includes the following steps: S401, photographing the fabric; S402, uploading the fabric photo to the cloud; S403, comparing the fabric photo with a reference photo; When the fabric product moves under the industrial camera, the industrial camera takes photos of the fabric product from multiple angles and directions. After the photos are taken, they are uploaded to the cloud. The photos taken in the cloud are compared one by one with the comparison photos taken previously to perform defect comparison detection on the fabric product.

[0010] Furthermore, the S104 further includes the following steps: S501. Compare and check whether there are any defects; S502, checking the type of fabric defects; Among them, during the comparative test, the fabric product photo is first compared with the intact fabric comparison photo, and the mutual comparison is used to determine whether the fabric product being tested has defects. When defects are detected, the fabric product being tested is compared with the fabric photo with the defective defect, and the type of defect of the fabric product being tested is detected by comparison, such as broken warp, broken weft and holes. If the defect type cannot be matched when comparing the defects, the fabric product being tested is compared with the intact fabric comparison photo again, and a new round of comparison is performed to prevent misjudgment. If the defect type is still not matched after multiple comparisons, it will be recorded through the cloud. If there is still no match after the re-comparison, the unmatched fabric will be recorded. The next day, the staff will re-place the unmatched fabric for testing and conduct manual comparison with the naked eye.

[0011] S503: The fabric being tested is intact; S504: Classify the comparative analysis results and record and store them in the cloud; S505, displaying the comparison and analysis results on the terminal; Among them, after the inspection is completed, the inspection results are classified into defective, non-defective and unmatched, and the classified analysis results are stored in the cloud. The measured data is displayed on the display terminal through the cloud for staff to view.

[0012] S506, mobile transmission fabric; After the detection is completed, the conveyor is restarted through the control circuit, and the conveyor is used to move and transport the detected fabric product.

[0013] Furthermore, the S105 further includes the following steps: S601, transmitting the fabric after inspection; S602, classifying and transmitting the fabric according to whether the fabric has defects; S603, the intact fabrics are moved and stacked; S604, the fabrics with defects are moved and stacked; S605, unmatched fabrics are moved and stacked; Among them, when transmitting fabric products, the conveyor adjusts the transmission direction according to the fabric products being inspected, and distinguishes defective, non-defective and unmatched fabric products based on the comparison results recorded in the cloud storage, and transmits and sorts them in different directions, classifying and transmitting the fabric products, and achieving sorting and classification during the transmission process.

[0014] Furthermore, the textile product defect detection system includes a shooting module, a mobile positioning module, a comparison module, an analysis module and a sorting module. The shooting module is composed of an industrial camera and a wireless transmitter, which is used to shoot intact fabrics and different types of defective fabrics as comparison references, and compare the photos of the inspected fabric products with the photos of the comparison references. The mobile positioning module is composed of a sorting conveyor, an infrared positioning sensor and a control circuit, which is used to move and transmit the inspected fabric products, and position the fabric products so that the fabric products move and stay under the industrial camera for photo comparison and detection. The comparison module is a cloud platform for storing photos taken by the industrial camera and performing comparison processing. The analysis module is a cloud platform for comparing the photos of the inspected fabric products with the photos of the comparison references and determining whether there are defects and the types of defects. The sorting module is a sorting conveyor for transmitting the inspected fabric products and classifying fabric products with defects and intact fabric products.

[0015] In summary, due to the adoption of the above technology, the beneficial effects of the present invention are: 1. During the comparative test, first compare the fabric product photo with the intact fabric comparison photo, and judge whether the fabric product under test has defects by mutual comparison. When defects are detected, the fabric product under test is compared with the fabric photo with defects, and the type of defects of the fabric product under test is determined by comparison, such as broken warp, broken weft and holes, etc. If the defect type cannot be matched when comparing the defects, the fabric product under test is compared with the intact fabric comparison photo again and the test is repeated. A round of comparison is performed to prevent misjudgment. If there is still no match with the defect type after multiple comparisons, it will be recorded through the cloud. If there is still no match after re-comparison, the unmatched fabric will be recorded. The next day, the staff will re-place the unmatched fabric for inspection, and conduct manual comparison with naked eye observation. The fabric product is photographed and compared with the reference fabric photo for inspection. The defect type of the fabric product is matched and confirmed and multiple rounds of inspection are carried out to prevent the shadows caused by indentations or wrinkles on the fabric product from affecting the inspection and comparison and causing misjudgment.

[0016] 2. Lay the fabric product flat on the conveyor, and use the conveyor to move the fabric. When the fabric product moves, the infrared positioning sensor is used to detect the area directly below. When the fabric product moves below the infrared positioning sensor, the infrared positioning sensor locates the fabric product. The infrared positioning sensor sends a signal and controls the conveyor to pause through the control circuit, so that the fabric product stays under the infrared positioning sensor and is aligned with the industrial camera. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 1 is a flowchart of steps S101 to S105 in a method and system for detecting defects in textile products according to the present invention; Figure 2 Schematic diagram of the flow of steps S201 to S203 in a method and system for detecting defects in textile products according to the present invention; Figure 3 Schematic diagram of the flow of steps S301 to S303 in a method and system for detecting defects in textile products according to the present invention; Figure 4 Schematic diagram of the flow of steps S401 to S403 in a method and system for detecting defects in textile products according to the present invention; Figure 5 Schematic diagram of the flow of steps S501 to S506 in a method and system for detecting defects in textile products according to the present invention; Figure 6 Schematic diagram of the flow of steps S601 to S605 in a method and system for detecting defects in textile products according to the present invention; Figure 7 This is a schematic diagram of the overall modules in a method and system for detecting defects in textile products according to the present invention.

[0018] In the figure: 1. Shooting module; 2. Mobile positioning module; 3. Comparison module; 4. Analysis module; 5. Sorting module. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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 ordinary technicians in this field without making creative work are within the scope of protection of the present invention. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the invention for which protection is sought, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0020] The present invention provides Figure 1 As shown, a method for detecting defects in textile products comprises the following steps: S101. Take reference pictures of intact fabrics; S102, moving the fabric product and aligning and positioning it; S103, comparing the fabric product with a reference picture; S104, analyzing and comparing whether the fabric has defects and the types of defects; S105, performing mobile sorting on the fabrics.

[0021] In actual use, step S101 is to smooth out and lay out intact fabric products and fabric products with different defects, i.e., level them. The fabric products are photographed from multiple directions using an industrial camera, etc., and the photographs are used as reference for comparison during inspection.

[0022] In the actual use process of step S102, the fabric product is laid on a conveyor, the fabric product is moved and transported by the conveyor, the fabric product is positioned, and the fabric product is inspected after being positioned.

[0023] In the actual application process, in step S103, after the fabric product is positioned, the fabric product is photographed by an industrial camera, and the photograph of the fabric product is compared with a previously taken reference photograph of an intact fabric product to detect whether the fabric product has any defects; The normalized cross correlation (NCC) algorithm is used to calculate the similarity between the tested fabric product photos and the reference photos.

[0024] In actual use, step S104 compares the fabric product with a photo of an intact fabric product to detect whether there are any defects on the fabric product. When a difference is detected between the fabric product and the photo of the intact fabric product, the fabric product is compared with a photo of the defective fabric product to determine the type of defect on the fabric product.

[0025] In the actual application process of step S105, after the inspection of the fabric products is completed, the fabric products are continuously transported by the conveyor. During transportation, the intact and defective fabric products are moved and sorted by conveyors in different transmission directions to classify and differentiate them.

[0026] In one implementation of this embodiment, Figure 2 As shown, step S101 further includes the following steps: S201, industrial camera takes pictures; S202, uploading the captured photos to the cloud; S203: Set the captured photo as a reference.

[0027] In the actual application process of steps S201-S203, intact fabric products are laid out flat, and the intact fabric products are photographed in multiple directions and angles by an industrial camera. After the shooting is completed, fabric products with different types of defects are laid out flat, and the defective fabric products are photographed in multiple directions and angles by an industrial camera. After the shooting is completed, the photographs are uploaded to the cloud and stored in the cloud as reference pictures. The stored photos are used for reference and comparison during inspection.

[0028] In one implementation of this embodiment, Figure 3 As shown, step S102 further includes the following steps: S301, placing the fabric on a conveyor; S302, conveying the fabric through a conveyor; S303: Positioning is performed using an infrared positioning sensor.

[0029] During the actual use of steps S301-S303, the fabric product is laid flat on a conveyor, and the conveyor drives the fabric to move and transport. When the fabric product moves, the infrared positioning sensor is used to detect the area directly below. When the fabric product moves below the infrared positioning sensor, the infrared positioning sensor positions the fabric product. The infrared positioning sensor sends a signal and controls the conveyor to pause through the control circuit, so that the fabric product stays under the infrared positioning sensor and is aligned with the industrial camera.

[0030] In one implementation of this embodiment, Figure 4 As shown, step S103 further includes the following steps: S401, photographing the fabric; S402, uploading the fabric photo to the cloud; S403. Compare the fabric photo with the reference photo.

[0031] In actual use, steps S401-S403 involve capturing photos of a fabric product from multiple angles and directions while it is positioned beneath the industrial camera. The photos are then uploaded to the cloud, where they are compared with previously taken comparison photos to detect defects. Multiple light sources are used to eliminate shadows during the photography process.

[0032] In one implementation of this embodiment, Figure 5 As shown, step S104 further includes the following steps: S501. Compare and check whether there are any defects; S502, checking the type of fabric defects; S503: The fabric being tested is intact; S504: Classify the comparative analysis results and record and store them in the cloud; S505, displaying the comparison and analysis results on the terminal; S506. Move the transmission fabric.

[0033] In the actual application process of step S501-step S502, during the comparative detection, the fabric product photo is first compared with the intact fabric comparison photo, and the mutual comparison is used to determine whether the fabric product being tested has defects. When defects are detected, the fabric product being tested is compared with the fabric photo with the defective defect, and the type of defect of the fabric product being tested is determined by comparison, such as broken warp, broken weft and hole, etc. If the defect type cannot be matched when comparing the defect, the fabric product being tested is compared with the intact fabric comparison photo again, and a new round of comparison is performed to prevent misjudgment. If multiple comparisons still do not match the defect type, the fabric product being tested is compared with the intact fabric comparison photo again, and a new round of comparison is performed to prevent misjudgment. If there is a type match, it will be recorded through the cloud. If there is still no match after re-comparison, the unmatched fabric will be recorded. The next day, the staff will re-place the unmatched fabric for inspection, and conduct manual comparison with naked eye observation. The fabric product will be photographed and compared with the reference fabric photo for inspection. The defect type of the fabric product will be matched and confirmed and multiple rounds of inspection will be carried out to prevent the shadows caused by indentations or wrinkles on the fabric product from affecting the inspection and comparison and causing misjudgment. The multiple comparisons are generally set to 3 times, that is, the comparison will be terminated after a maximum of 3 comparisons. The SVM classifier or convolutional neural network (CNN) model is used to judge the defect type, such as broken warp and broken weft, based on texture features such as LBP and HOG.

[0034] In the actual application process of steps S503-S505, after the detection is completed, the detection results are classified into defective, non-defective and unmatched, and the classified analysis results are stored in the cloud. The measured data is displayed on the display terminal through the cloud for staff to view.

[0035] In step S506, during actual use, after the detection is completed, the conveyor is restarted through the control circuit to move and transport the inspected fabric product through the conveyor.

[0036] In one implementation of this embodiment, Figure 6 As shown, step S105 further includes the following steps: S601, transmitting the fabric after inspection; S602, classifying and transmitting the fabric according to whether the fabric has defects; S603, the intact fabrics are moved and stacked; S604, the fabrics with defects are moved and stacked; S605. The unmatched fabrics are moved and stacked; that is, the unmatched fabrics are transferred to a manual re-inspection station in real time to avoid production interruption.

[0037] During the actual use of steps S601-S605, when transmitting fabric products, the conveyor adjusts the transmission direction according to the inspected fabric products, and distinguishes defective, non-defective, and unmatched fabric products based on the comparison results recorded in the cloud storage, and transmits and sorts them in different directions, classifying and transmitting the fabric products, and achieving sorting and classification during the transmission process.

[0038] The textile product defect detection system includes a shooting module 1, a mobile positioning module 2, a comparison module 3, an analysis module 4 and a sorting module 5. The shooting module 1 is composed of an industrial camera and a wireless transmitter, which is used to shoot intact fabrics and different types of defective fabrics as comparison references, and compare the photos of the inspected fabric products with the photos of the comparison references. The mobile positioning module 2 is composed of a sorting conveyor, an infrared positioning sensor and a control circuit, which is used to move and transmit the inspected fabric products, and position the fabric products so that the fabric products move and stay under the industrial camera for photo comparison and detection. The comparison module 3 is a cloud platform for storing photos taken by the industrial camera and performing comparison processing. The analysis module 4 is a cloud platform for comparing the photos of the inspected fabric products with the photos of the comparison references and determining whether there are defects and the types of defects. The sorting module 5 is a sorting conveyor for transmitting the inspected fabric products and classifying fabric products with defects and intact fabric products.

[0039] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A method for detecting defects in textile products, comprising the following steps: S101. Take reference pictures of intact fabrics; Among them, intact fabric products and fabric products with different defects are smoothed and laid out, and the fabric products are photographed from multiple directions using industrial cameras, and the photographs are used as reference for comparison during inspection; S102, moving the fabric product and aligning and positioning it; The fabric product is placed on a conveyor, the fabric product is moved and transported by the conveyor, the fabric product is positioned, and the fabric product is inspected after being positioned; S103, comparing the fabric product with a reference picture; After the fabric product is positioned, it is photographed by an industrial camera and compared with a previously taken reference photo of an intact fabric product to detect whether the fabric product has any defects. S104, analyzing and comparing whether the fabric has defects and the types of defects; When comparing fabric products, first compare them with photos of intact fabric products to detect whether there are any defects on the fabric products. If a difference is detected between the fabric products and the intact fabric products, then compare them with photos of fabric products with defects to determine the type of defects on the fabric products. S105, mobile sorting of fabrics; Among them, after the inspection of the fabric products is completed, the fabric products are further transported through the conveyor. During transportation, the intact and defective fabric products are moved and sorted by conveyors with different transmission directions to classify and distinguish them.

2. The method according to claim 1, wherein: The S101 further includes the following steps: S201, industrial camera takes pictures; S202, uploading the captured photos to the cloud; S203, setting the captured photo as a reference; Among them, the intact fabric products are laid out flat, and the intact fabric products are photographed in multiple directions and angles by industrial cameras. After the shooting is completed, the fabric products with different types of defects are laid out flat, and the defective fabric products are photographed in multiple directions and angles by industrial cameras. After the shooting is completed, the photos are uploaded to the cloud and stored in the cloud as reference pictures. The stored photos are used for reference and comparison during inspection.

3. The method according to claim 1, wherein: The S102 further includes the following steps: S301, placing the fabric on a conveyor; S302, conveying the fabric through a conveyor; S303, positioning through an infrared positioning sensor; Among them, the fabric product is laid flat on the conveyor, and the conveyor drives the fabric to move and transport. When the fabric product moves, the infrared positioning sensor is used to detect the area directly below. When the fabric product moves under the infrared positioning sensor, the infrared positioning sensor positions the fabric product. The infrared positioning sensor sends a signal and controls the conveyor to pause through the control circuit, so that the fabric product stays under the infrared positioning sensor and is aligned with the industrial camera.

4. The method according to claim 1, wherein: The S103 further includes the following steps: S401, photographing the fabric; S402, uploading the fabric photo to the cloud; S403, comparing the fabric photo with a reference photo; Among them, when the fabric product moves under the industrial camera, the industrial camera takes photos of the fabric product from multiple angles and directions. After the shooting is completed, the photos are uploaded to the cloud. The photos taken in the cloud are compared one by one with the comparison photos taken previously to perform defect comparison detection on the fabric product.

5. The method according to claim 1, wherein: The S104 further includes the following steps: S501. Compare and check whether there are any defects; S502, checking the type of fabric defects; Among them, during the comparative test, the fabric product photo is first compared with the intact fabric comparison photo, and the mutual comparison is used to determine whether the fabric product being tested has defects. When defects are detected, the fabric product being tested is compared with the fabric photo with the defective defect, and the type of defect of the fabric product being tested is determined by comparison, such as broken warp, broken weft and hole. If the defect type cannot be matched when comparing the defects, the fabric product being tested is compared with the intact fabric comparison photo again, and a new round of comparison is performed to prevent misjudgment. If the defect type is still not matched after multiple comparisons, it will be recorded through the cloud. If there is still no match after the re-comparison, the unmatched fabric will be recorded. The next day, the staff will re-place the unmatched fabric for testing and perform manual comparison with the naked eye. S503: The fabric being tested is intact; S504: Classify the comparative analysis results and record and store them in the cloud; S505, displaying the comparison and analysis results on the terminal; After the test is completed, the test results are classified into defective, non-defective and unmatched, and the classified analysis results are stored in the cloud. The measured data is displayed on the display terminal through the cloud for staff to view; S506, mobile transmission fabric; After the detection is completed, the conveyor is restarted through the control circuit, and the conveyor is used to move and transport the detected fabric product.

6. The method according to claim 1, wherein: The S105 further includes the following steps: S601, transmitting the fabric after inspection; S602, classifying and transmitting the fabric according to whether the fabric has defects; S603, the intact fabrics are moved and stacked; S604, the fabrics with defects are moved and stacked; S605, unmatched fabrics are moved and stacked; Among them, when transmitting fabric products, the conveyor adjusts the transmission direction according to the fabric products being inspected, and distinguishes defective, non-defective and unmatched fabric products based on the comparison results recorded in the cloud storage, and transmits and sorts them in different directions, classifying and transmitting the fabric products, and achieving sorting and classification during the transmission process.

7. A textile product defect detection system, characterized by: It includes a shooting module, a mobile positioning module, a comparison module, an analysis module and a sorting module. The shooting module is composed of an industrial camera and a wireless transmitter, which is used to shoot intact fabrics and different types of defective fabrics as comparison references, and to compare and detect the photos of the inspected fabric products with the photos of the comparison references. The mobile positioning module is composed of a sorting conveyor, an infrared positioning sensor and a control circuit, which is used to move and transmit the inspected fabric products, and position the fabric products so that the fabric products move and stay under the industrial camera for photo comparison and detection. The comparison module is a cloud platform for storing photos taken by the industrial camera and performing comparison processing. The analysis module is a cloud platform for comparing the photos of the inspected fabric products with the photos of the comparison references and determining whether there are defects and the types of defects. The sorting module is a sorting conveyor for transmitting the inspected fabric products and classifying fabric products with defects and intact fabric products.

Images