A real-time color difference calculation method and device for textiles based on hyperspectral

Abstract

The present invention relates to the technical field of textile detection, and particularly relates to a real-time color difference calculation method and device for textiles based on hyperspectrum. Among them, a real-time color difference calculation method for textiles based on hyperspectrum includes the following steps: S1: Initialize the real-time color difference calculation device for textiles; S2: Drive the textile to be measured through the support part of the real-time color difference calculation device for textiles to keep it flat, and drive the textile to pass through the spectral acquisition station in sequence; S3: The original spectral data of the textile is obtained by spectral acquisition through the hyperspectral acquisition module preset in the real-time color difference calculation device for textiles, and the total real-time color difference of the textile is calculated and output. The present invention can realize a series of tasks such as the support and automatic leveling of textiles, the automatic conveying of textiles, the automatic spectral acquisition and color difference calculation of textiles, etc. It has a high degree of automation and intelligence, greatly reduces the labor intensity of operators, and can effectively improve the efficiency of textile detection.

Description

Technical Field

[0001] The present invention relates to the technical field of textile detection, and particularly relates to a method and device for real-time color difference calculation of textiles based on hyperspectrum. Background Art

[0002] The uniformity of textile dyeing and the color difference between batches have always been key problems that have attracted much attention and are urgently to be solved in the textile production process. At present, most of the color difference inspections for textiles adopt traditional manual sampling and evaluation methods. However, the efficiency of manual sampling and evaluation is low, and due to the differences in individual color sensitivity and inconsistent habits of each person, the judgment result standards are not unified. At the same time, due to the inconsistent silk thread density of textiles, different tensile strengths have a great impact on the test results, and the differences in manual detection methods and detection angles are likely to lead to misjudgment of good products, causing unnecessary losses to manufacturers. In the prior art, there are also devices and apparatuses that apply spectrometers to the field of color detection. However, there are still many limitations when applied to flexible materials such as textiles. The existing color difference detection devices cannot perform automatic color difference detection work. Only operators can manually use the color difference detection device to sample textile samples, and its detection effect is low, and there is also the problem of missing areas, which is likely to lead to the outflow of defective products. At the same time, the existing detection methods also have great technical difficulties in detecting textiles such as stockings that are thin, elastic, knitted sparsely, and have large holes. When elastic fabrics are placed on common automated detection devices such as conveyor belts, wrinkles will occur, resulting in inaccurate measurement. And thin and semi-transparent textiles have high light transmittance, and the reflected spectral signals are weak, resulting in greater interference from transmitted light and other reflected light to the results. Moreover, the silk thread density of textiles is different, and for relatively sparse textile samples, the holes will also have a great impact on the sample test results, and the color difference detection accuracy is low. Summary of the Invention

[0003] The technical problem to be solved by the present invention: The existing textile detection methods have a large proportion of manual participation, the detection environment cannot be guaranteed to be unified, and there are many external environmental interferences, resulting in poor textile detection effects.

[0004] To solve the above technical problems, the first aspect of the present invention adopts the following technical solution: A method for real-time color difference calculation of textiles based on hyperspectrum, comprising the following steps:

[0005] S1: Initialize the real-time color difference calculation device for textiles, and install the textile on the support part of the real-time color difference calculation device for textiles;

[0006] S2: Drive the textile to be measured to move through the support part of the real-time color difference calculation device for textiles to keep it flat, and drive the textile to pass through the spectral acquisition station in turn;

[0007] S3: The hyperspectral acquisition module preset in the real-time textile color difference calculation device acquires the original spectral data of the textile through spectral acquisition, and calculates and outputs the total real-time color difference of the textile.

[0008] When the present invention works, it can realize a series of operations such as the support and automatic leveling of textiles, the automatic conveying of textiles, the automatic spectral acquisition and color difference calculation of textiles, with high automation and intelligence levels, greatly reducing the labor intensity of operators, effectively improving the efficiency of textile detection, unifying the detection environment at the same time, being able to avoid the influence of external factors on the detection results, improving the detection accuracy, driving the textile to be measured to move through the support part of the real-time textile color difference calculation device to keep it flat, being applicable to the color difference detection of flexible textiles, and being able to avoid the influence of wrinkles and holes during excessive stretching on the detection results.

[0009] Preferably, in the step S2, when driving the textile to be measured to move through the support part of the real-time textile color difference calculation device to keep it flat, the following steps are adopted: obtaining the specification parameters of the textile, and driving the textile to be measured to move through changing the size of the support part of the real-time textile color difference calculation device to keep it flat.

[0010] Preferably, in the step S2, when driving the textile to be measured to move through the support part of the real-time textile color difference calculation device to keep it flat, the following steps are adopted: continuously collecting the depth information of the textile through the depth acquisition module preset in the real-time textile color difference calculation device to obtain the real-time depth information of the textile, judging whether the textile is flat by calculating the depth information, and when the textile is not flat, changing the size of the support part of the real-time textile color difference calculation device to drive the textile to be measured to move to keep it flat.

[0011] Preferably, in the step S2, when continuously collecting the depth information of the textile through the depth acquisition module preset in the real-time textile color difference calculation device to obtain the real-time depth information of the textile, judging whether the textile is flat by calculating the depth information, and when the textile is not flat, changing the size of the support part of the real-time textile color difference calculation device to drive the textile to be measured to move to keep it flat, the following steps are adopted:

[0012] A1: Obtaining the standard depth value obtained from the depth acquisition module and the textile thickness;

[0013] A2: The depth acquisition module collects the depth information of the textile entering its detection area to obtain the depth information of each point;

[0014] A3: Subtracting the depth information of each point from the standard depth value, comparing the absolute value of the difference with a preset depth threshold, and marking the points greater than the depth threshold as abnormal points;

[0015] A4: Count the number of abnormal points. When the proportion of abnormal points is greater than or equal to the preset flatness threshold, determine that the current state of the textile is uneven, and change the size of the support part of the real-time color difference calculation device of the textile to drive the textile to move.

[0016] When the present invention works, it can dynamically adjust the size of the support part according to the specification parameters of the textile, is applicable to the detection work of textiles of different specifications, has a wide application range and good versatility. At the same time, through the way that the depth acquisition module preset by the real-time color difference calculation device of the textile continuously collects the depth information of the textile, the real-time monitoring and real-time adjustment of the flatness of the textile can be realized, and the accuracy of textile detection can be further improved by keeping the textile flat.

[0017] Preferably, in the step S2, when driving the textile to pass through the spectral acquisition station in sequence, the following steps are adopted: connect the head end of the textile with the preset pulling belt, and the winding part of the textile winding mechanism preset by the real-time color difference calculation device of the textile is connected with the textile through the pulling belt to drive the textile to pass through the spectral acquisition station at a uniform speed. The transfer speed of the textile is set by the following formula:

[0018] ;

[0019] Where: v is the transfer speed of the textile, x is the slit width, H is the distance between the acquisition part of the hyperspectral acquisition module and the spectral acquisition station, f is the focal length of the hyperspectral acquisition module, and fps is the frame rate of the hyperspectral acquisition module.

[0020] When the present invention works, by driving the textile to pass through the spectral acquisition station at a uniform speed according to the preset transfer speed, the transfer speed of the textile can be adapted to the frame rate of the hyperspectral acquisition module, and while maintaining the acquisition effect and not missing the sample measurement, the detection efficiency can be guaranteed.

[0021] Preferably, in the step S3, when the hyperspectral acquisition module preset by the real-time color difference calculation device of the textile obtains the original spectral data of the textile through spectral acquisition and calculates and outputs the total real-time color difference of the textile, the following steps are adopted:

[0022] B1: Remove the interfering light in the spectral acquisition station through the light absorption module preset by the real-time color difference calculation device of the textile, and obtain the original spectral data of the textile through spectral acquisition by the hyperspectral acquisition module preset by the real-time color difference calculation device of the textile;

[0023] B2: Calculate the reflectivity through the original spectral data of the textile, and calculate several groups of LAB data from the reflectivity;

[0024] B3: For several groups of LAB data in the original spectral data of textiles, perform regional averaging operations to obtain a set of LAB data after regional averaging. After calculating the LAB values of this set of LAB data, calculate and output the real-time total color difference of the textiles.

[0025] When the present invention works, the hyperspectral acquisition module obtains the original spectral data of textiles through spectral acquisition, can implement measurement and batch calculation of the LAB values and total color differences of textiles, and can further calculate the color difference values that meet international standards, which is convenient for classifying and rating textiles. At the same time, through regional averaging operations, it can avoid the interference of textile holes on the detection results, has strong anti-interference ability, and can ensure the accuracy of the detection results.

[0026] To solve the above technical problems, the second aspect of the present invention adopts the following technical solution: A real-time color difference calculation device for textiles based on hyperspectrum, applying a real-time color difference calculation method for textiles based on hyperspectrum as described above, including:

[0027] A light source module, used to provide a light source with adjustable angle and uniform irradiation;

[0028] A depth acquisition module, used to acquire color information and depth information in the spectral acquisition station;

[0029] A hyperspectral acquisition module, used to acquire spectral data in the spectral acquisition station;

[0030] A textile support mechanism, used to support textiles of different specifications according to the acquired depth information to keep them flat;

[0031] A textile winding mechanism, used to transfer textiles through the spectral acquisition station and wind up the textiles after detection;

[0032] A light absorption module, used to avoid interference light in the spectral acquisition station;

[0033] The light-emitting part of the light source module, the acquisition part of the depth acquisition module, and the acquisition part of the hyperspectral acquisition module are all optically connected to the spectral acquisition station. The textile support mechanism and the textile winding mechanism are respectively arranged on both sides of the spectral acquisition station. The light absorption part of the light absorption module is arranged on the periphery of the spectral acquisition station. The support part of the textile support mechanism acts after the depth acquisition module acquires the depth information of the textile to drive the textile to keep flat. The speed of the textile winding mechanism for transferring textiles is set to be synchronized with the acquisition work of the hyperspectral acquisition module.

[0034] When the present invention is working, it can realize a series of tasks such as supporting and automatically leveling textiles, automatically conveying textiles, automatically collecting spectra of textiles and calculating color differences, etc. It has a high degree of automation and intelligence, greatly reduces the labor intensity of operators, and can effectively improve the efficiency of textile detection. At the same time, the light absorption module is used to avoid interfering light in the spectrum collection station, and the influence of external factors on the detection results can be avoided, thereby improving the detection accuracy. The depth collection module and the textile support mechanism are used to drive the movement of the textile to be tested to keep it flat. It is suitable for color difference detection of flexible textiles, and can avoid the influence of wrinkles and holes caused by excessive stretching on the detection results.

[0035] Preferably, the textile support mechanism comprises at least two support rods, a transverse slide rail and a driving device, wherein the transverse slide rail extends in a direction perpendicular to the moving direction of the textile, and the support rods are slidably mounted on the transverse slide rail, and adjacent support rods are kept parallel, and each support rod is connected to the driving device through a corresponding transmission assembly, and the spacing between adjacent support rods is adjusted under the drive of the driving device to flatten the supported textile.

[0036] Preferably, the depth acquisition module collects depth information of the textile when the textile moves into the spectrum acquisition station to obtain depth information of several points on the textile, and the spacing between adjacent support rods is adjusted by comparing the depth information of each point with a preset standard depth value.

[0037] Preferably, the textile winding mechanism comprises a textile driving mechanism for uniformly transferring the textile at a preset transfer speed and a textile storage mechanism for storing the inspected textiles, the textile driving mechanism being provided with two driving wheels and a driven wheel, the two driving wheels and the driven wheel being arranged in parallel, a gap for the textiles to pass through being provided between the driven wheel and the two driving wheels, the driven wheel being moved after the textiles are loaded to press the textiles against the two driving wheels respectively, the rotation directions of the two driving wheels being arranged in opposite directions, the feeding end of the textile driving mechanism being connected to the discharging end of the textile supporting mechanism, and the discharging end of the textile driving mechanism being connected to the storage part of the textile storage mechanism.

[0038] The beneficial technical effects of the present invention include:

[0039] 1. The present invention can realize a series of operations such as the support and automatic leveling of textiles, the automatic conveying of textiles, the automatic spectral acquisition of textiles, and the color difference calculation. It has a high degree of automation and intelligence, greatly reducing the labor intensity of operators, effectively improving the efficiency of textile detection. At the same time, it unifies the detection environment, can avoid the influence of external factors on the detection results, improves the detection accuracy, and drives the textile to be measured to move through the support part of the real-time color difference calculation device of the textile to keep it flat, which is suitable for the color difference detection of flexible textiles and can avoid the influence of wrinkles and holes during excessive stretching on the detection results.

[0040] 2. The present invention can dynamically adjust the size of the support part according to the specification parameters of the textile, which is suitable for the detection work of textiles with different specifications, has a wide application range and good versatility. At the same time, by continuously collecting the depth information of the textile through the depth acquisition module preset in the real-time color difference calculation device of the textile, the real-time monitoring and real-time adjustment of the flatness of the textile can be realized, and the detection accuracy of the textile can be further improved by keeping the textile flat.

[0041] 3. The present invention drives the textile to pass through the spectral acquisition station at a constant speed according to the preset transfer speed, which can make the transfer speed of the textile match the frame frequency of the hyperspectral acquisition module, maintain the acquisition effect, ensure the detection efficiency without missing samples.

[0042] 4. The present invention uses a hyperspectral acquisition module to obtain the original spectral data of the textile through spectral acquisition, can measure and batch calculate the LAB value and total color difference of the textile, and can further calculate the color difference value that meets international standards, which is convenient for classifying and rating the textiles. At the same time, through the regional averaging operation, the interference of the holes in the textile on the detection results can be avoided, and the anti-interference ability is strong, which can ensure the accuracy of the detection results.

[0043] 5. The present invention can realize a series of operations such as the support and automatic leveling of textiles, the automatic conveying of textiles, the automatic spectral acquisition of textiles, and the color difference calculation. It has a high degree of automation and intelligence, greatly reducing the labor intensity of operators, effectively improving the efficiency of textile detection. At the same time, through the light absorption module, the interference light in the spectral acquisition station can be avoided, the influence of external factors on the detection results can be avoided, and the detection accuracy is improved. Through the depth acquisition module and the textile support mechanism, the textile to be measured is driven to move to keep it flat, which is suitable for the color difference detection of flexible textiles and can avoid the influence of wrinkles and holes during excessive stretching on the detection results.

[0044] Other features and advantages of the present invention will be disclosed in detail in the following specific embodiments and drawings. Brief Description of the Drawings

[0045] The present invention will be further described below in conjunction with the accompanying drawings:

[0046] Figure 1 It is a flowchart of the working process of a real-time color difference calculation method for textiles based on hyperspectral.

[0047] Figure 2 It is a flowchart of the working process of step S2 in a real-time color difference calculation method for textiles based on hyperspectral.

[0048] Figure 3 It is a flowchart of the working process of step S3 in a real-time color difference calculation method for textiles based on hyperspectral.

[0049] Figure 4 It is a schematic structural diagram of a real-time color difference calculation device for textiles based on hyperspectral.

[0050] Figure 5 It is a schematic structural diagram of a textile support mechanism. Specific embodiments

[0051] The technical solutions of the embodiments of the present invention will be explained and described below in conjunction with the accompanying drawings of the embodiments of the present invention. However, the following embodiments are only the preferred embodiments of the present invention, not all of them. Based on the embodiments in the embodiments, other embodiments obtained by those skilled in the art without creative work all fall within the protection scope of the present invention.

[0052] In the following description, terms such as "inner", "outer", "upper", "lower", "left", "right", etc. indicating orientation or position relationship are only for convenience of describing the embodiments and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation to the present invention. Embodiment 1:

[0053] Please refer to Figure 1 , this embodiment discloses a real-time color difference calculation method for textiles based on hyperspectral, including the following steps:

[0054] S1: Initialize the real-time color difference calculation device for textiles, and install the textile on the support part of the real-time color difference calculation device for textiles;

[0055] S2: Drive the textile to be measured to move through the support part of the real-time color difference calculation device for textiles to keep it flat, and drive the textile to pass through the spectral acquisition station in turn;

[0056] S3: The original spectral data of the textile is obtained by spectral acquisition through the hyperspectral acquisition module 3 preset in the real-time color difference calculation device for textiles, and the real-time total color difference of the textile is calculated and output to realize the comprehensive detection of the textile.

[0057] When this embodiment works, it can achieve a series of operations such as the support and automatic leveling of textiles, the automatic conveying of textiles, the automatic spectral collection and color difference calculation of textiles, with high automation and intelligence levels, greatly reducing the labor intensity of operators, effectively improving the efficiency of textile detection, unifying the detection environment at the same time, being able to avoid the influence of external factors on the detection results, improving the detection accuracy, driving the textile to be measured to move through the support part of the real-time textile color difference calculation device to keep it flat, being applicable to the color difference detection of flexible textiles, and being able to avoid the influence of wrinkles and holes during excessive stretching on the detection results.

[0058] In this embodiment, in step S2, when driving the textile to be measured to move through the support part of the real-time textile color difference calculation device to keep it flat, the following steps are adopted: obtaining the specification parameters of the textile, and driving the textile to be measured to move through changing the size of the support part of the real-time textile color difference calculation device to keep it flat.

[0059] Preferably, in step S2, when driving the textile to be measured to move through the support part of the real-time textile color difference calculation device to keep it flat, the following steps are adopted: continuously collecting the depth information of the textile through the preset depth collection module 2 of the real-time textile color difference calculation device to obtain the real-time depth information of the textile, judging whether the textile is flat by calculating the depth information, and when the textile is not flat, changing the size of the support part of the real-time textile color difference calculation device to drive the textile to be measured to move to keep it flat.

[0060] Please refer to Figure 2 , in specific implementation, in step S2, when continuously collecting the depth information of the textile through the preset depth collection module 2 of the real-time textile color difference calculation device to obtain the real-time depth information of the textile, judging whether the textile is flat by calculating the depth information, and when the textile is not flat, changing the size of the support part of the real-time textile color difference calculation device to drive the textile to be measured to move to keep it flat, the following steps are adopted:

[0061] A1: Obtain the standard depth value obtained from the depth collection module 2 and the textile thickness;

[0062] A2: The depth collection module 2 collects the depth information of the textile entering its detection area to obtain the depth information of each point;

[0063] A3: Subtract the depth information of each point from the standard depth value, take the absolute value of the difference and compare it with the preset depth threshold, and mark the points greater than the depth threshold as abnormal points;

[0064] A4: Count the number of abnormal points. When the proportion of abnormal points is greater than or equal to the preset flatness threshold, determine that the current state of the textile is uneven, and change the size of the support part of the real-time color difference calculation device of the textile to drive the textile to move.

[0065] During the operation of this embodiment, the size of the support part can be dynamically adjusted according to the specification parameters of the textile, which is suitable for the detection of textiles with different specifications, has a wide application range and good versatility. At the same time, through the way that the depth acquisition module 2 preset in the real-time color difference calculation device of the textile continuously acquires the depth information of the textile, the real-time monitoring and real-time adjustment of the flatness of the textile can be realized, and the accuracy of textile detection can be further improved by keeping the textile flat.

[0066] As a further improvement of this embodiment, in step S2, when driving the textile to pass through the spectral acquisition station in turn, the following steps are adopted: connect the leading end of the textile with the preset pulling belt, and the winding part of the textile winding mechanism 5 preset in the real-time color difference calculation device of the textile is connected with the textile through the pulling belt to drive the textile to pass through the spectral acquisition station at a constant speed. The transfer speed of the textile is set by the following formula:

[0067] ;

[0068] Where: v is the transfer speed of the textile, x is the slit width, H is the distance between the acquisition part of the hyperspectral acquisition module 3 and the spectral acquisition station, f is the focal length of the hyperspectral acquisition module 3, and fps is the frame rate of the hyperspectral acquisition module 3.

[0069] During the operation of this embodiment, by driving the textile to pass through the spectral acquisition station at a constant speed according to the preset transfer speed, the transfer speed of the textile can be adapted to the frame rate of the hyperspectral acquisition module 3, while maintaining the acquisition effect and ensuring the detection efficiency without missing samples.

[0070] Preferably, in step S2, the following steps are further included: place a correction circle on the textile, drive the correction circle to move by the textile, and obtain the ratio of the diameter in the displacement direction of the correction circle to the diameter in the direction perpendicular to the displacement direction to correct the transfer speed of the textile, which can realize further adjustment of the transfer speed, facilitate the operator to make adaptive adjustment according to the actual situation, and thus ensure the detection accuracy. Embodiment 2:

[0071] Please refer to Figure 3 , this embodiment provides a method for calculating the real-time color difference of textiles based on hyperspectrum. The same parts as other embodiments will not be described in detail. The differences will be described in detail below.

[0072] In this embodiment, in step S3, when the total color difference of textiles is calculated in real time by the preset hyperspectral acquisition module 3 of the textile real-time color difference calculation device and output after calculation, the following steps are adopted:

[0073] B1: Remove the interfering light in the spectral acquisition station through the preset light absorption module 6 of the textile real-time color difference calculation device, and obtain the original spectral data of the textile by spectral acquisition through the preset hyperspectral acquisition module 3 of the textile real-time color difference calculation device;

[0074] B2: Calculate the reflectance from the original spectral data of the textile, and obtain several groups of LAB data from the reflectance;

[0075] B3: Obtain a set of LAB data after regional averaging for several groups of LAB data in the original spectral data of the textile. After calculating the LAB value of this set of LAB data, calculate and output the total color difference of the textile in real time.

[0076] When this embodiment works, the preset hyperspectral acquisition module 3 is used to obtain the original spectral data of the textile by spectral acquisition, which can implement the measurement and batch calculation of the LAB value and total color difference of the textile, and can further calculate the color difference value that meets the international standard, facilitating the classification and grading of textiles. At the same time, through the regional averaging operation, the interference of textile holes on the detection results can be avoided, with strong anti-interference ability, and the accuracy of the detection results can be guaranteed.

[0077] As a further improvement of this embodiment, in step B1, the following steps are also included: collect dark background data and whiteboard data, and correct the original spectral data of the textile and the whiteboard data through the dark background data;

[0078] In step B1, when correcting the original spectral data of the textile and the whiteboard data through the dark background data, the following calculation formula is adopted:

[0079] ;

[0080] ;

[0081] Where: is the dark background data, is the collected whiteboard data, is the whiteboard data after black field correction, is the collected original spectral data of the textile, is the original spectral data of the textile after black field correction;

[0082] In step B2, when calculating the reflectance from the original spectral data of the textile, the following calculation formula is adopted:

[0083] ;

[0084] Wherein: is the reflectance.

[0085] Preferably, in step B2, when calculating several groups of LAB data from the reflectance, the following steps are adopted: calculate the current tristimulus values from the reflectance, and call the tristimulus values of a perfect reflecting diffuser under the standard illuminant to calculate the LAB values, and repeatedly obtain several groups of LAB data in turn.

[0086] In step B2, when calculating the current tristimulus values from the reflectance, the following calculation formula is adopted:

[0087] ;

[0088] ;

[0089] ;

[0090] Wherein: X, Y, and Z are the current tristimulus values, is the relative spectral power distribution of the standard illuminant, , , are the spectral tristimulus values of the CIE standard observer, is the wavelength interval, and k is the normalization constant.

[0091] In step B2, when calculating the LAB values by calling the tristimulus values of a perfect reflecting diffuser under the standard illuminant, the following calculation formula is adopted:

[0092] ;

[0093] ;

[0094] ;

[0095] Wherein: , , are the LAB values, , , are the tristimulus values of a perfect reflecting diffuser under the standard illuminant, is the color space conversion function. In this embodiment, can be defined as the following formula:

[0096] ;

[0097] Wherein: takes the value .

[0098] Preferably, in step B3, when obtaining a set of LAB data after regional averaging from several sets of LAB data in the original spectral data of the textile through regional averaging operation, the following steps are adopted: Set the average number of pixels as n. For several sets of LAB data, select a pixel as the center, and take n pixels above, below, to the left, and to the right of it, obtaining a total of pixel points. For pixel points, perform an average calculation to obtain the value of the center point. Use as the window size of the sliding window, slide it from left to right in sequence, slide the corresponding number of pixels each time according to the preset step size, slide from top to bottom, and slide the corresponding number of pixels each time according to the preset step size until all the pixels of several sets of LAB data are slid, obtaining a new set of LAB data. During operation, the value of n needs to be selected according to the actual specification parameters of the textile. For example, for stockings with different needle counts, for stockings with 0 to 300 needle counts, the holes are relatively large. At this time, the value of n can be taken as 7, which can avoid the influence of holes on the detection results to a certain extent. In the needle count range of 301 to 512, the value of n can be taken as 5. In the needle count range of 513 to 720, the value of n can be taken as 3. When the needle count is above 721, there are no holes interfering with the color difference detection in the stockings. At this time, the value of n can be taken as 1.

[0099] In step B3, when obtaining a set of LAB data after regional averaging from several sets of LAB data in the original spectral data of the textile through regional averaging operation and calculating the LAB value of this set of LAB data, the following steps are adopted: For pixels that meet the condition that when serving as the central pixel, the number of surrounding pixels plus the number of its own pixels meets , update the LAB value after regional averaging using the following formula:

[0100] ;

[0101] ;

[0102] ;

[0103] Among them: , , are the LAB values after regional averaging, , , are the LAB values before regional averaging, , , is the average value of pixel points;

[0104] For pixels that do not meet the condition, update the LAB value after regional averaging using the following formula:

[0105] ;

[0106] ;

[0107] ;

[0108] In step B3, when calculating the total real-time color difference of the textile, the following calculation formula is adopted:

[0109] ;

[0110] Where: is the total color difference of the textile, , , are respectively the differences in , , values between the current color and the standard color. Preferably, the total color difference can also be compared with the threshold value in the GSC grade standard, and the GSC grade of the textile can be determined according to the comparison result. At the same time, the total color difference can also be calculated in real time by any existing standard calculation method such as CIE2000 or CMC. Embodiment 3:

[0111] This embodiment provides a real-time color difference calculation method for textiles based on hyperspectral. The same parts as other embodiments will not be described in detail, and the differences will be described in detail below.

[0112] In this embodiment, in order to further improve the textile detection efficiency, it is also possible to detect the defect of the textile color difference while detecting the defect of the textile flaw, which greatly improves the production efficiency and reduces the outflow of defective products. At the same time, for elastic textiles, they have a certain retraction stress, and some relatively concealed flaw defects or flaw defects such as needle path defects and missed stitch defects are also more difficult to detect under the action of the retraction stress. In this embodiment, in the step S2, the following steps are further included: obtaining the specification parameters of the textile, and driving the textile to be measured to move by changing the size of the end of the support part of the real-time color difference calculation device for textiles according to the preset stretching ratio table to stretch it to the appropriate specification and keep it flat. In specific implementation, the stretching ratio of the textile needs to be custom-selected according to the specification of the textile to be measured, so as to achieve that the textile located at the first end of the textile transfer route in the spectral acquisition station remains flat and does not stretch, and stretch the textile located at the end of the textile transfer path in the spectral acquisition station to the appropriate specification and keep it flat.

[0113] In step S3, the following steps are further included. After calculating and outputting the total textile color difference of the textile located at the position near the head end of the textile transfer route in the spectral acquisition station, the total textile color difference of this part of the textile in the stretched state and the non-stretched state is output synchronously. By comparing whether the total textile color difference exceeds a preset threshold, it is determined whether the textile has a defect. During operation, stretching the textile can enlarge the holes in the textile. At this time, in the original spectral data of the textile, the textile area is the effective area, and the total textile color difference of this part is calculated during calculation. When the textile has a defect, there will be a large deviation in the total textile color difference of the textile in the stretched state and the non-stretched state, and this deviation will show an accelerating growth trend according to the change of the stretching ratio. By selecting an appropriate stretching ratio, the accuracy of textile defect detection can be further improved, and the outflow of defective products can be avoided. When calculating the total textile color difference of the textile in the stretched state and the non-stretched state, the following formula is used:

[0114] ;

[0115] ;

[0116] ;

[0117] ;

[0118] Wherein: is the total textile color difference of the textile in the stretched state and the non-stretched state, k is a correction coefficient, and in specific implementation, it needs to be selected according to the actual hairiness value of the textile, and the general selection range is 0.4 to 1, which can avoid misjudgment of holes caused by flying feathers;

[0119] , and are respectively the differences in LAB values between the current color and the color of the non-stretched textile;

[0120] , and are the LAB values of the textile in the stretched state;

[0121] , and are the LAB values of the textile in the non-stretched state. Embodiment 4:

[0122] Please refer to Figure 4 , this embodiment provides a device for calculating the real-time color difference of textiles based on hyperspectrum, applying a method for calculating the real-time color difference of textiles based on hyperspectrum as described in the above embodiment, including:

[0123] A light source module 1 for providing a light source with adjustable angle and uniform illumination;

[0124] A depth acquisition module 2 for acquiring color information and depth information within the spectral acquisition station. During operation, a ToF 3D camera can be used, and of course, any other suitable depth acquisition camera can also be used;

[0125] A hyperspectral acquisition module 3 for acquiring spectral data within the spectral acquisition station;

[0126] A textile support mechanism 4 for supporting textiles of different specifications according to the acquired depth information to keep them flat;

[0127] A textile winding mechanism 5 for transporting textiles through the spectral acquisition station and winding the inspected textiles;

[0128] A light absorption module 6 for preventing interfering light from appearing within the spectral acquisition station;

[0129] The light-emitting part of the light source module 1, the acquisition part of the depth acquisition module 2, and the acquisition part of the hyperspectral acquisition module 3 are all optically connected to the spectral acquisition station. The textile support mechanism 4 and the textile winding mechanism 5 are respectively arranged on both sides of the spectral acquisition station, and the light absorption part of the light absorption module 6 is arranged on the periphery of the spectral acquisition station. Preferably, a textile real-time color difference calculation device can also be deployed in a darkroom environment to avoid the interference of stray light in the overlapping band of external natural light and the hyperspectral acquisition module 3. The support part of the textile support mechanism 4 operates after the depth acquisition module 2 acquires the depth information of the textile to drive the textile to remain flat. The speed of the textile winding mechanism 5 for transporting textiles is set to be synchronized with the acquisition work of the hyperspectral acquisition module 3.

[0130] During the operation of this embodiment, a series of operations such as the support and automatic leveling of textiles, the automatic transportation of textiles, the automatic spectral acquisition and color difference calculation of textiles can be realized. The degree of automation and intelligence is high, greatly reducing the labor intensity of operators, effectively improving the efficiency of textile detection. At the same time, the light absorption module 6 prevents interfering light from appearing within the spectral acquisition station, avoiding the influence of external factors on the detection results and improving the detection accuracy. By driving the textile to be measured to move through the depth acquisition module 2 and the textile support mechanism 4 to keep it flat, it is suitable for the color difference detection of flexible textiles, and can avoid the influence of wrinkles and holes during overstretching on the detection results.

[0131] Please refer to Figure 5In a specific implementation, the textile support mechanism 4 includes at least two support rods 41, a transverse slide rail 42 and a driving device 43. The transverse slide rail 42 is extended and arranged in a direction perpendicular to the moving direction of the textile. The support rods 41 can be slidably limited and installed on the transverse slide rail 42. Adjacent support rods 41 remain parallel. Each support rod 41 is connected to the driving device 43 through a corresponding transmission assembly. The spacing between adjacent support rods 41 is adjusted under the drive of the driving device 43 to flatten the supported textile. Preferably, the depth acquisition module 2 collects the depth information of the textile when the textile moves to the spectrum acquisition station to obtain the depth information of several points on the textile. The spacing between adjacent support rods 41 is adjusted by comparing the depth information of each point with the preset standard depth value.

[0132] During operation, the connection portion between the material storage portion and the material discharge portion of the support rod 41 can be set to an S shape, which can prevent the textiles installed on the support rod 41 from being pulled out completely, thereby ensuring the uniformity of the material discharge. Preferably, a light-absorbing material can also be sprayed on the support rod 41 to avoid interference with reflected light caused by the reflection of the support rod 41. Preferably, it also includes a horizontal support module 8, which is arranged on the side of the spectrum collection station close to the textile winding mechanism 5, and the supporting portion of the horizontal support module 8 is arranged below the textile transfer path, so that the material discharge portion of the support rod 41 is kept flush with the supporting portion of the horizontal support module 8, which can prevent the textiles from vibrating during transfer and ensure the flatness of the textiles.

[0133] Preferably, the textile winding mechanism 5 includes a textile driving mechanism 51 for uniformly transferring the textile at a preset transfer speed and a textile storage mechanism 52 for storing the inspected textiles. The textile driving mechanism 51 is provided with two driving wheels 511 and a driven wheel 512. The two driving wheels 511 and the driven wheel 512 are arranged in parallel. There is a gap between the driven wheel 512 and the two driving wheels 511 for the textiles to pass through. The driven wheel 512 moves after the textiles are loaded to press the textiles against the two driving wheels 511 respectively. The rotation directions of the two driving wheels 511 are set to be opposite. The feeding end of the textile driving mechanism 51 is connected to the discharging end of the textile supporting mechanism 4, and the discharging end of the textile driving mechanism 51 is connected to the storage part of the textile storage mechanism 52.

[0134] Preferably, it also includes an angle adjustment mechanism 7, and the hyperspectral acquisition module 3 is installed on the rotating part of the angle adjustment mechanism 7. When working, the acquisition angle of the hyperspectral acquisition module 3 can be adjusted by the angle adjustment mechanism 7, such as 10° angle observation and 2° angle observation, etc., which can simulate the observation angle of the human eye and facilitate the unification of the working conditions of color difference detection.

[0135] The beneficial technical effects of this embodiment include: The present invention can achieve a series of operations such as the support and automatic leveling of textiles, the automatic conveying of textiles, the automatic spectral collection of textiles, and color difference calculation. It has a high degree of automation and intelligence, greatly reducing the labor intensity of operators, effectively improving the efficiency of textile detection. At the same time, it unifies the detection environment, can avoid the influence of external factors on the detection results, improves the detection accuracy, drives the textile to be measured to move through the support part of the real-time textile color difference calculation device to keep it flat, is suitable for the color difference detection of flexible textiles, and can avoid the influence of wrinkles and holes during overstretching on the detection results.

[0136] As described above, the above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Those skilled in the art should understand that the present invention includes not only the content described in the drawings and the above specific implementation manner. Any modification that does not deviate from the functional and structural principles of the present invention will be included in the scope of the claims.

Claims

1. A real-time color difference calculation method for textiles based on hyperspectral, characterized in that, It includes the following steps: S1: Initialize the textile real-time color difference calculation device, and install the textile on the supporting part of the textile real-time color difference calculation device; S2: Drive the textile to be measured to move through the support part of the textile real-time color difference calculation device to keep it flat, and drive the textile to pass through the spectral acquisition station in sequence; In the step S2, the following steps are further included: Obtain the specification parameters of the textile, and drive the textile to be measured to move according to the preset stretching ratio table by changing the size of the end of the support part of the textile real-time color difference calculation device to stretch it to the appropriate specification and keep it flat; S3: The original spectral data of the textile is obtained by spectral acquisition through the hyperspectral acquisition module (3) preset in the textile real-time color difference calculation device, and the real-time total color difference of the textile is calculated and output; In the step S3, the following steps are further included: After calculating and outputting the total color difference of the textile located at the position near the head end of the textile transfer route in the spectral acquisition station, synchronously output the total color difference of this part of the textile in the stretched state and the non-stretched state, and judge whether the textile has defect by comparing whether the total color difference of this part of the textile in the stretched state and the non-stretched state exceeds the preset threshold; When calculating the total color difference of the textile in the stretched state and the non-stretched state, the following formula is used: ； ； ； ； Wherein: is the total color difference of the textile in the stretched state and the non-stretched state, and k is a correction coefficient; , and are respectively the differences in LAB values between the color of the textile in the stretched state and the color of the textile when not stretched. , and are the LAB values of the textile in the stretched state. , and are the LAB values of the textile in the unstretched state.

2. The real-time color difference calculation method of textiles based on hyperspectral according to claim 1, characterized in that: In the step S2, when driving the textile to be measured to move through the support part of the textile real-time color difference calculation device to keep it flat, the following steps are adopted: Obtain the specification parameters of the textile, and drive the textile to be measured to move to keep it flat by changing the size of the support part of the textile real-time color difference calculation device; 3. A real-time color difference calculation method for textiles based on hyperspectral as claimed in claim 1, characterized in that: In the step S2, when driving the textile to be measured to move through the support part of the textile real-time color difference calculation device to keep it flat, the following steps are adopted: Continuously collect the depth information of the textile through the depth acquisition module (2) preset in the textile real-time color difference calculation device to obtain the real-time depth information of the textile, calculate the depth information to judge whether the textile is flat, and when the textile is not flat, change the size of the support part of the textile real-time color difference calculation device to drive the textile to be measured to move to keep it flat; 4. A real-time color difference calculation method for textiles based on hyperspectral according to claim 3, characterized in that: In the step S2, when continuously collecting the depth information of the textile through the depth acquisition module (2) preset in the textile real-time color difference calculation device to obtain the real-time depth information of the textile, calculating the depth information to judge whether the textile is flat, and when the textile is not flat, changing the size of the support part of the textile real-time color difference calculation device to drive the textile to be measured to move to keep it flat, the following steps are adopted: A1: Obtain the standard depth value obtained from the depth acquisition module (2) and the textile thickness; A2: The depth acquisition module (2) collects the depth information of the textile entering its detection area to obtain the depth information of each point; A3: Subtract the depth information of each point from the standard depth value, take the absolute value of the difference and compare it with the preset depth threshold, and mark the points greater than the depth threshold as abnormal points; A4: Count the number of abnormal points. When the proportion of abnormal points is greater than or equal to the preset flatness threshold, determine that the current state of the textile is uneven, and change the size of the support part of the real-time color difference calculation device of the textile to drive the textile to move.

5. A real-time color difference calculation method for textiles based on hyperspectral, as claimed in claim 1, wherein: In the step S2, when driving the textile to pass through the spectral acquisition station in sequence, the following steps are adopted: connect the head end of the textile with the preset pulling belt, and connect the winding part of the textile winding mechanism (5) preset by the real-time color difference calculation device of the textile with the textile through the pulling belt, and drive the textile to pass through the spectral acquisition station at a constant speed. The transfer speed of the textile is set by the following formula: ； Where: v is the transfer speed of the textile, x is the slit width, H is the distance from the acquisition part of the hyperspectral acquisition module (3) to the spectral acquisition station, f is the focal length of the hyperspectral acquisition module (3), and fps is the frame rate of the hyperspectral acquisition module (3).

6. The real-time color difference calculation method for textiles based on hyperspectral according to claim 1, characterized in that: In the step S3, when the original spectral data of the textile is obtained by spectral acquisition through the hyperspectral acquisition module (3) preset by the real-time color difference calculation device of the textile and the total real-time color difference of the textile is calculated and output, the following steps are adopted: B1: Remove the interference light in the spectral acquisition station through the light absorption module (6) preset by the real-time color difference calculation device of the textile, and obtain the original spectral data of the textile by spectral acquisition through the hyperspectral acquisition module (3) preset by the real-time color difference calculation device of the textile; B2: Calculate the reflectivity through the original spectral data of the textile, and calculate several groups of LAB data from the reflectivity; B3: Obtain a set of LAB data after regional averaging for several groups of LAB data in the original spectral data of the textile, calculate the LAB value of this set of LAB data, and then calculate and output the total real-time color difference of the textile.

7. A real-time textile color difference calculation device based on hyperspectral, applying a real-time textile color difference calculation method based on hyperspectral as described in any one of claims 1 to 6, characterized in that, Including: A light source module (1) for providing a light source with adjustable angle and uniform illumination; A depth acquisition module (2) for acquiring color information and depth information in the spectral acquisition station; A hyperspectral acquisition module (3) for acquiring spectral data in the spectral acquisition station; A textile support mechanism (4) for supporting textiles of different specifications according to the acquired depth information to keep them flat; A textile winding mechanism (5) for moving the textile through the spectral acquisition station and winding the tested textile; A light absorption module (6) for preventing interference light from appearing in the spectral acquisition station; The light-emitting part of the light source module (1), the acquisition part of the depth acquisition module (2), and the acquisition part of the hyperspectral acquisition module (3) are all optically connected to the spectral acquisition station. The textile support mechanism (4) and the textile winding mechanism (5) are respectively arranged on both sides of the spectral acquisition station. The light absorption part of the light absorption module (6) is arranged on the periphery of the spectral acquisition station. The support part of the textile support mechanism (4) moves after the depth acquisition module (2) acquires the depth information of the textile to drive the textile to be flat. The transfer speed of the textile winding mechanism (5) for moving the textile is set to be synchronized with the acquisition work of the hyperspectral acquisition module (3).

8. The real-time color difference calculation device for textiles based on hyperspectral according to claim 7, characterized in that: The textile support mechanism (4) includes at least two support rods (41), a transverse sliding rail (42) and a driving device (43). The transverse sliding rail (42) extends along a direction perpendicular to the moving direction of the textile. The support rods (41) are slidably and limitably mounted on the transverse sliding rail (42), and adjacent support rods (41) are kept parallel. Each support rod (41) is in transmission connection with the driving device (43) through a corresponding transmission component. The distance between adjacent support rods (41) is adjusted under the drive of the driving device (43) to flatten the supported textile.

9. The real-time color difference calculation device for textiles based on hyperspectral according to claim 8, characterized in that: When the textile is displaced into the spectral acquisition station, the depth acquisition module (2) acquires the depth information of the textile to obtain the depth information of several points on the textile. The distance between adjacent support rods (41) is adjusted by comparing the depth information of each point with a preset standard depth value.

10. A real-time color difference calculation device for textiles based on hyperspectral, as claimed in claim 7, characterized in that: The textile winding mechanism (5) includes a textile driving mechanism (51) for uniformly transferring the textile at a preset transfer speed and a textile storage mechanism (52) for storing the textile after the detection. The textile driving mechanism (51) is provided with two driving wheels (511) and a driven wheel (512). The two driving wheels (511) and the driven wheel (512) are arranged in parallel. A gap for the textile to pass through is provided between the driven wheel (512) and each of the two driving wheels (511). After the textile is loaded, the driven wheel (512) acts to press the textile against the two driving wheels (511) respectively. The rotation directions of the two driving wheels (511) are set to be opposite. The feeding end of the textile driving mechanism (51) is connected to the discharging end of the textile support mechanism (4), and the discharging end of the textile driving mechanism (51) is connected to the storage part of the textile storage mechanism (52).

Images