Rapid lossless agilawood authenticity identification method based on microscopic imaging and electronic nose

Abstract

The invention provides an agilawood rapid lossless authenticity identification method based on microscopic imaging and an electronic nose, the method is based on an agilawood authenticity identification model of microscopic images and an agilawood authenticity identification model of smell, and the method comprises the following steps: S1, adopting a microscope and an imaging system to obtain an agilawood detection sample microscopic structure image; s2, inputting the microscopic structure image of the agilawood detection sample into an agilawood authenticity identification model based on the microscopic image; s3, according to the smell-based agilawood authenticity identification model, obtaining data of the agilawood detection sample in each sensor of the electronic nose; s4, starting an automatic mode recognition system of the electronic nose, and recognizing data of the smells of the electronic nose in 40s, 50s and 60s of the agilawood detection; the method disclosed by the invention is simple, convenient and efficient, is simple to operate, does not need to pre-treat a sample, solves the problems of subjective influence and poor repeatability during traditional artificial evaluation, and greatly improves the authenticity identification precision of the agilawood.

Description

Technical Field

[0001] The invention relates to the technical field of authenticity identification and quality detection of precious wood, and in particular to a method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and an electronic nose. Background Art

[0002] As a traditional Chinese medicine, agarwood has antibacterial, anti-inflammatory, antitussive, cardiovascular disease relief, gastrointestinal protection, and sleep promotion effects. As a spice, the main components of agarwood's secondary metabolites include sesquiterpenes and chromones, which are the material basis for agarwood's medicinal effects and aroma.

[0003] Existing identification technologies mainly include property identification, microscopic identification (tissue structure characteristics), physical and chemical identification (color reaction, thin layer chromatography, liquid chromatography, etc.), etc., among which the latter two methods are feasible, but have the disadvantages of cumbersome operation, long time, high cost, etc., and the agarwood test sample needs to be destroyed, and the rapid and non-destructive identification of the authenticity of agarwood cannot be achieved; property identification is to identify through sensory methods such as "eyes, hands, noses, and mouths", but as a subjective sensory evaluation method, it lacks objectivity and cannot be objectively quantified. The evaluation of the color, feel, and smell of agarwood varies from person to person, and the accuracy of authenticity identification is unstable. With the increase in the trading volume of agarwood raw materials, beads, handicrafts, etc. in the market, it is urgent to find a method that is fast, non-destructive, accurate, and can objectively and quantitatively identify the authenticity of agarwood.

[0004] CN 118781368A discloses a method for identifying agarwood by image acquisition. On the one hand, the clarity of microscopic image acquisition is reduced for samples with uneven surfaces; on the other hand, since the samples of agarwood bracelets are formed by curve rotary cutting, the tube holes and wood ray characteristics are deformed, making it difficult to obtain the real agarwood tube hole characteristic image and difficult to compare with the traditional agarwood characteristic image. Moreover, the large agarwood sold on the market has oily overflow characteristics due to its rich oil content, which is easy to be confused with fake agarwood such as oil injection. Most importantly, many fake agarwood samples on the market currently use microbubble oil, micro-oil injection and other methods, and the microscopic structure has not changed significantly, making it difficult to identify through microscopic images. CN114112951A invented a method for high-spectrum identification of insect-leaked agarwood, and CN.105699343A proposed a fluorescence spectral imaging technology, which identifies by testing the spectral characteristics of the sample and processing the data. However, due to the complex composition of agarwood, the data processing mode analysis is slow, which is not suitable for the market's requirements for rapid detection of beads.

[0005] At present, counterfeiting of agarwood mainly includes impersonation of non-Aquilaria wood, or adding agarwood extract or other oils to inferior agarwood by high pressure, micro pressure, oil pressing, drug soaking, high polishing, etc. In these counterfeiting methods, the appearance or fragrance of agarwood changes, and a single identification method is difficult to identify all counterfeiting methods. For example, microscopic imaging can identify samples of non-Aquilaria wood impersonation, high polishing, high pressure, and drug soaking through surface color changes, distribution and width of the internal phloem, etc., while samples counterfeited by soaking in agarwood extract, micro pressure, etc. are difficult to identify. The electronic nose is a device that simulates the biological olfactory system. It uses an embedded gas sensor array to detect and identify various complex odors. When the odor molecules contact the sensor array and react chemically, a series of odor fingerprints will be obtained. The different odors of genuine and fake agarwood will produce specific responses. These responses can be accurately identified as fake agarwood with different odors through pattern recognition algorithms. Therefore, in view of the market sales demand for agarwood raw materials, beads, and small handicrafts in different forms, it is urgent to develop a highly accurate, widely applicable, fast, and non-destructive agarwood identification method. Summary of the invention

[0006] To achieve the above object, the present invention provides a method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose, wherein the method is based on an agarwood authenticity identification model based on microscopic images and an agarwood authenticity identification model based on odor, and comprises the following steps:

[0007] S1: Use microscope and imaging system to obtain microstructure images of agarwood samples;

[0008] S2: Input the microstructure image of the agarwood test sample into the agarwood authenticity identification model based on microscopic images. If the judgment result is ≥90%, it is authentic agarwood, and the test is stopped. If the judgment result is <90%, the authenticity identification based on odor analysis is performed;

[0009] S3: According to the agarwood authenticity identification model based on smell, the data of the agarwood test sample in each sensor of the electronic nose is obtained;

[0010] S4: Turn on the automated pattern recognition system of the electronic nose, input the electronic nose odor recognition data of 40s, 50s, and 60s of agarwood detection into the clustering recognition model of the authenticity sample, and use the Euclidean algorithm, the deterministic finite automatic algorithm, and the segmented expansion algorithm to give the discrimination results. If all three algorithms are judged to be true, the detection sample is judged to be authentic agarwood. If one or more algorithms are judged to be false, the detection sample is judged to be counterfeit agarwood.

[0011] Optionally, the agarwood authenticity identification model based on microscopic images comprises the following steps:

[0012] S11: Use microscope and imaging system to obtain microstructure images of authentic agarwood samples;

[0013] S12: extracting features from the image data, extracting features from the sample microstructure image, and obtaining features such as pores, internal phloem, wood fibers, or wood rays of the sample, as well as features such as surface finish, color, oil distribution, and oil overflow;

[0014] S13: Obtain the microstructure image feature vector of the authentic and fake agarwood samples; train the sample feature vector using a set algorithm to obtain an agarwood recognition model.

[0015] Optionally, the agarwood authenticity identification model based on smell comprises the following steps:

[0016] S31: Place the agarwood sample in a sealed container and leave it at room temperature for 30 to 60 minutes, or shake the sample at 60°C for 10 to 20 minutes;

[0017] S32: Turn on the electronic nose, preheat the device for 20 to 30 minutes, insert the injection needle of the sensor into the soft cover of the sealed injection bottle, the distance between the injection needle and the bottle mouth is 1 / 3 of the bottle height, and then insert the air inlet needle of the sensor into the soft cover of the injection bottle, the distance between the injection needle and the bottle mouth is greater than the distance between the injection needle and the bottle mouth by 1mm; then set the array parameters of the sensor for detection: gas flow rate 0.5-0.7L / min, cleaning time 50-100s, detection time 80-100s, start sampling and measurement at 15-30℃, and obtain the data of the authenticity detection sample in each sensor of the electronic nose respectively;

[0018] S33: Turn on the automatic pattern recognition system of the electronic nose, and use linear discriminant analysis to cluster the electronic nose gas recognition data of 40s, 50s, and 60s of the authentic and fake agarwood samples, respectively, to obtain a cluster recognition model for the authentic and fake agarwood samples.

[0019] Optionally, the imaging system includes a lower rod, and a first-axis gimbal connected to the upper end of the lower rod, the free end of the first-axis gimbal is connected to the second-axis gimbal, the second-axis gimbal is in a "U"-shaped structure, an imaging body is rotatably arranged in the second-axis gimbal, a sensor is arranged on the lower end surface of the imaging body, and also includes a height lifting mechanism for connecting the first-axis gimbal and the second-axis gimbal. When working, the sensor collects a vertical distance signal between the lower end of the imaging body and the collected object, and adjusts the position of the imaging body in the vertical direction through the height lifting mechanism according to the vertical distance signal to adjust the vertical distance between the lower end of the imaging body and the collected object.

[0020] Optionally, the height lifting mechanism includes a connecting component connected to the upper end surface of the second-axis gimbal, a first lifting rod is fixedly connected to the upper end surface of the connecting component, a supporting ring is provided on an outer movable sleeve of the first lifting rod, a first fixing plate is fixedly connected to the outer side wall of the supporting ring, the free end of the first fixing plate is fixedly connected to the second fixing plate, the second fixing plate is fixedly arranged on the upper end surface of the first-axis gimbal, the upper end of the first lifting rod is fixedly connected to the first fixing block, the side wall of the first fixing block is fixedly connected to the third fixing plate, the free end of the third fixing plate is fixedly connected to the second fixing block, the lower end surface of the second fixing block is fixedly connected to the second lifting rod, the lower end of the second lifting rod is fixedly connected to the driving end of the driving cylinder, and the driving cylinder is fixedly arranged on the first-axis gimbal, wherein the first-axis gimbal and the second-axis gimbal are staggered in the vertical direction.

[0021] Optionally, the connecting assembly includes a mounting base fixedly connected to the second-axis gimbal, the mounting base is slidably connected to the first-axis gimbal, a first rotating rod is fixedly connected to the upper end surface of the mounting base, the upper end of the first rotating rod is fixedly connected to the driving end of the driving motor, and the driving motor is fixedly arranged on the lower end of the first lifting rod.

[0022] Optionally, the imaging system also includes a position locking mechanism for locking the mounting base and the first-axis pan-tilt head to each other, the position locking mechanism including a plurality of position locking grooves provided on the mounting base, a position locking rod being provided in one of the position locking grooves, the position locking rod being movably provided in a first movable groove, the first movable groove being provided on the first-axis pan-tilt head, one end of the position locking rod located in the first movable groove being fixedly connected to a first magnet block, the first magnet block being movably inserted in the first movable groove, and a first electromagnet matching the first magnet block being fixedly connected to a side wall of the first movable groove.

[0023] Optionally, the position locking mechanism also includes a first guide support assembly arranged between the first magnet block and the first electromagnet and used to connect the first magnet block and the first electromagnet, the first guide support assembly includes a first guide support rod, a first guide support tube and a first guide support spring, the first guide support rod is movably inserted in the first guide support tube, the first guide support spring is wound outside the first guide support rod, and the two ends of the first guide support spring are respectively fixedly connected to the side wall of the first guide support rod and the outer side wall of the first guide support tube.

[0024] Optionally, the imaging system also includes a fastening mechanism, which includes a rigid fastening plate and an elastic fastening plate connected to each other, the elastic fastening plate is arranged close to the imaging body, the rigid fastening plate and the elastic fastening plate are movably arranged in a second movable groove, the second movable groove is opened on the second axis pan-tilt platform, a first movable rod is fixedly connected to the side wall of the rigid fastening plate, the first movable rod is movably inserted into the second axis pan-tilt platform, a second magnet block is fixedly connected to the free end of the first movable rod, a second electromagnet matching the second magnet block is fixedly connected to the side wall of the second axis pan-tilt platform, the second electromagnet is movably sleeved outside the first movable rod, a second guide support assembly is fixedly connected to the side wall of the second magnet block, the free end of the second guide support assembly is fixedly connected to the second axis pan-tilt platform, the second guide support assembly includes a second guide support rod, a second guide support cylinder and a second guide support spring, the second guide support rod is movably inserted into the second guide support cylinder, the second guide support spring is wound and connected outside the second guide support rod, and the two ends of the second guide support spring are respectively fixedly connected to the side wall of the second guide support rod and the outer side wall of the second guide support cylinder.

[0025] Optionally, the imaging system also includes a heat dissipation mechanism, which includes a connecting plate connected to the second magnet block, the side wall of the connecting plate is fixedly connected to a second moving rod, the second moving rod is fixedly sleeved with a driving screw rod, the driving screw rod is movably plugged into the second-axis gimbal, the driving screw rod is externally threaded with an internally threaded tube, the internally threaded tube is rotatably set on the side wall of the second-axis gimbal, the internally threaded tube is fixedly sleeved with a first driving gear, the first driving gear and the second driving gear are meshed with each other, the second driving gear is fixedly sleeved outside the second rotating rod, one end of the second rotating rod is rotatably set on the side wall of the second-axis gimbal, and the other end of the second rotating rod is fixedly connected to a rotating fan.

[0026] The beneficial effects of the present invention are as follows:

[0027] The method of the present invention is simple, efficient, easy to operate, and does not require pre-treatment of the sample. It also overcomes the problems of subjective influence and poor repeatability in traditional manual evaluation, greatly improving the accuracy of authenticity identification of agarwood. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is the microscopic structural image of the authentic agarwood in the present invention;

[0029] Figure 2 This is a microscopic structural image of the oil-filled agarwood in the present invention;

[0030] Figure 3 This is the microscopic structure image of the high-polished agarwood in the present invention;

[0031] Figure 4 This is a microscopic structural image of the fake agarwood that is not agarwood in the present invention;

[0032] Figure 5 This is a schematic diagram of clustering genuine and fake agarwood in the odor analysis system of the present invention;

[0033] Figure 6 This is a diagram showing the discrimination results of agarwood detection samples based on the odor analysis model in the present invention;

[0034] Figure 7 It is a structural schematic diagram of an existing imaging system;

[0035] Figure 8 It is a schematic structural diagram of the imaging system in the method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose of the present invention;

[0036] Fig. 9 The invention discloses a rapid and non-destructive method for distinguishing the authenticity of agarwood based on microscopic imaging and electronic nose. Figure 8 A is a magnified schematic diagram of the results;

[0037] Fig.10 It is a schematic diagram of the structure of the connection components in the method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose of the present invention;

[0038] Fig.11 The invention discloses a rapid and non-destructive method for distinguishing the authenticity of agarwood based on microscopic imaging and electronic nose. Fig. 9 A schematic diagram of the partial top-down cross-sectional structure in the middle;

[0039] Fig.12 It is a structural schematic diagram of the fastening mechanism in the method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose of the present invention;

[0040] Fig.13 The present invention is a schematic diagram of the structure of the heat dissipation mechanism in the method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose.

[0041] Reference numerals

[0042] Lower rod 1, first axis gimbal 2, second axis gimbal 3, imaging body 4, sensor 5, height lifting mechanism 6, connecting assembly 61, driving motor 611, first rotating rod 612, mounting base 613, first lifting rod 62, supporting ring 63, first fixing plate 64, second fixing plate 65, first fixing block 66, third fixing plate 67, second fixing block 68, second lifting rod 69, driving cylinder 610, position locking mechanism 7, position locking groove 71, position locking rod 72, first moving groove 73, first magnet block 74, first guide support assembly 75, first guide support rod 751, first guide support cylinder 752, first guide support spring 753, first electromagnet 76,

[0043] The fastening mechanism 8, the rigid fastening plate 81, the elastic fastening plate 82, the second moving groove 83, the first moving rod 84, the second magnet block 85, the second electromagnet 86, the second guide support assembly 87, the second guide support rod 871, the second guide support cylinder 872, the second guide support spring 873,

[0044] Heat dissipation mechanism 9 , connecting plate 91 , second moving rod 92 , driving screw rod 93 , internal threaded tube 94 , first driving gear 95 , second driving gear 96 , second rotating rod 97 , rotating fan 98 . DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solutions and advantages 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 of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein should be understood by people with general skills in the field to which the present invention belongs. "Including" and similar words used in this article mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0046] In view of the problems existing in the prior art, the embodiments of the present invention provide a method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose. Figures 1 to 6 As shown, the method is based on a microscopic image-based agarwood authenticity identification model and an odor-based agarwood authenticity identification model, comprising the following steps:

[0047] S1: Use microscope and imaging system to obtain microstructure images of agarwood samples;

[0048] S2: Input the microstructure image of the agarwood test sample into the agarwood authenticity identification model based on microscopic images. If the judgment result is ≥90%, it is authentic agarwood, and the test is stopped. If the judgment result is <90%, the authenticity identification based on odor analysis is performed;

[0049] S3: According to the agarwood authenticity identification model based on smell, the data of the agarwood test sample in each sensor of the electronic nose is obtained;

[0050] S4: Turn on the automated pattern recognition system of the electronic nose, input the electronic nose odor recognition data of 40s, 50s, and 60s of the agarwood detection into the clustering recognition model of the authenticity sample, and use the Euclidean algorithm, the deterministic finite automatic algorithm, and the segmented expansion algorithm to give the discrimination results. If all three algorithms are judged to be true, the detection sample is judged to be authentic agarwood. If one or more algorithms are judged to be false, the detection sample is judged to be counterfeit agarwood.

[0051] Or use other types of sensor-type electronic noses to obtain the test sample odor data, and input the electronic nose odor recognition data of 40s, 50s, and 60s of agarwood detection into the clustering recognition model of the true and false samples, and use the Euclidean algorithm (Euklid), the deterministic finite automatic algorithm (DFA) and the piecewise expansion algorithm (KOR) to give the discrimination results. If all three algorithms are judged to be true, the test sample is judged to be authentic agarwood. If one or more algorithms are judged to be false, the test sample is judged to be counterfeit agarwood.

[0052] In one embodiment, the agarwood authenticity identification model based on microscopic images comprises the following steps:

[0053] S11: Use microscope and imaging system to obtain microstructure images of authentic agarwood samples;

[0054] S12: extracting features from the image data, extracting features from the sample microstructure image, and obtaining features such as pores, internal phloem, wood fibers, or wood rays of the sample, as well as features such as surface finish, color, oil distribution, and oil overflow;

[0055] S13: Obtain the microstructure image feature vector of the authentic and fake agarwood samples; train the sample feature vector using a set algorithm to obtain an agarwood recognition model.

[0056] In one embodiment, the agarwood authenticity identification model based on smell comprises the following steps:

[0057] S31: Place the agarwood sample in a sealed container and leave it at room temperature for 30 to 60 minutes, or shake the sample at 60°C for 10 to 20 minutes;

[0058] S32: Turn on the electronic nose (the model used in the electronic nose is PEN3), preheat the equipment for 20 to 30 minutes, insert the injection needle of the sensor into the soft cover of the sealed injection bottle, the distance between the needle tip of the injection needle and the bottle mouth is 1 / 3 of the bottle height, and then insert the air inlet needle of the sensor into the soft cover of the injection bottle, the distance between the needle tip of the air inlet needle and the bottle mouth is greater than the distance between the needle tip of the injection needle and the bottle mouth by 1mm; then set the array parameters of the sensor for detection: gas flow rate 0.5-0.7L / min, cleaning time 50-100s, detection time 80-100s, start sampling and determination at 15-30℃, and obtain the data of the authenticity detection sample in each sensor of the electronic nose respectively; or use other types of sensor-type electronic noses, set the array parameters of the sensor for detection;

[0059] S33: Turn on the automatic pattern recognition system of the electronic nose, and use linear discriminant analysis to cluster the electronic nose gas identification data of 40s, 50s, and 60s of the authentic and fake agarwood samples, respectively, to obtain a clustering recognition model for the authentic and fake agarwood samples; or use the odor data of the authentic and fake agarwood samples obtained by other types of sensor-type electronic noses, and use linear discriminant analysis (LDA) to cluster the electronic nose odor identification data of 40s, 50s, and 60s of the authentic and fake agarwood samples, respectively, to obtain a clustering recognition model for the authentic and fake agarwood samples.

[0060] The feature extraction of the agarwood sample image data to obtain the sample feature vector includes determining the coordinate values ​​of the sample pores, axial parenchyma, or wood rays in the coordinate system corresponding to the structural image according to the pores, axial parenchyma, or wood ray features;

[0061] Determine the coordinate value of the sample foreign body in the coordinate system corresponding to the structure image according to the surface finish feature;

[0062] Determine the coordinate value of the sample spilled grease in the coordinate system corresponding to the constructed image according to the grease spill feature;

[0063] A digital matrix is ​​constructed according to the determined coordinate values, and the digital matrix is ​​determined as the feature vector of the authentic agarwood sample.

[0064] The feature extraction module includes: a first extraction submodule, which is used to extract features from the sample microstructure image to obtain the sample pore, axial parenchyma, or wood ray features; a second extraction submodule, which is used to extract features from the sample microstructure image to obtain the sample surface finish features; and a third extraction submodule, which is used to extract features from the sample's spilled grease to obtain the sample's grease spill features.

[0065] The method for establishing the agarwood identification model based on microscopic images is as follows:

[0066] (1) taking a string of genuine agarwood beads, scanning each bead using the microscopic imaging system of the present invention, obtaining a microscopic structural image of the genuine agarwood, and storing the image in a computer in the form of a picture;

[0067] (2) Take a string of fake agarwood beads, use the microscopic imaging system of the present invention to scan each bead, obtain a microscopic structural image of the fake agarwood, and store it in a computer in the form of a picture, such as Figure 2 As shown;

[0068] (3) The original image is preprocessed by edge segmentation and threshold segmentation to obtain clear image features of authentic and fake agarwood. Based on the grayscale co-occurrence matrix and differential statistical extraction, the features of the microscopic structure images of authentic and fake agarwood samples, such as pores, internal phloem, wood fibers, or wood rays, as well as surface smoothness, color, oil distribution, oil overflow and other features closely related to the authenticity of agarwood are extracted. The image data is then annotated to establish an image digital matrix of authentic and fake agarwood collectibles.

[0069] (4) Matlab software was used to extract and analyze image features, and the feature images were processed and further divided into multiple dimensional data sets for training and testing of agarwood identification models.

[0070] (5) The features of authentic agarwood images are automatically extracted and identified through a deep convolutional neural network algorithm. The convolutional neural network includes at least an input layer, a convolution layer, a pooling layer, and an output layer.

[0071] (6) The sample feature vector is trained based on the set algorithm, and the processed image is randomly divided into training data, verification data and test data in a ratio of 8:1:1. The Softmax loss function is used to train the network model to form an agarwood authenticity recognition model.

[0072] (7) An agarwood authenticity identification module, which is used to use the agarwood authenticity identification model to identify the agarwood sample image to be identified, and obtain the authenticity identification result of the agarwood to be identified.

[0073] A method for rapid and lossless authenticity identification of agarwood based on microscopic imaging and electronic nose comprises the following steps:

[0074] Construction of the microscopic imaging system: It consists of a microscope, an imaging system, a sensor, and a storage card.

[0075] Placement of agarwood test samples based on microscopic inspection: Place the agarwood test samples on the stage under the microscope, adjust the height of the stage, and for beading, the direction of the beading can be adjusted through the motion control system.

[0076] Imaging of agarwood samples based on microscopic testing: adjust the magnification by 20 times, move the stage height until the image is clear and the structural features are obvious, and save the image.

[0077] Image preprocessing: Gray-level co-occurrence matrix and differential statistics are used to extract features such as pores, internal phloem, wood fibers, or wood rays in the microstructure image, as well as surface smoothness, color, oil distribution, oil overflow and other features closely related to the authenticity of agarwood, to obtain feature images.

[0078] Authenticity identification: Input the feature image into the agarwood authenticity identification module, match it with the feature quantities in the authentic and counterfeit agarwood database, and determine the authenticity by similarity.

[0079] The parameters that need to be adjusted in the microscopic imaging system include: position and magnification.

[0080] Smell system construction: It consists of an electronic nose, a sealed container, and a water bath oscillating pot. The sealed container should be less than 500mL, and the agarwood sample and the test sample should be completely placed in the sealed container. According to the volume of the agarwood sample and the test sample, the smallest container that can be placed should be selected. Generally, 20mL, 50mL, 100mL, 300mL, and 500mL sealed containers are selected.

[0081] Odor analysis: Based on the agarwood authenticity identification model of microscopic images, the agarwood test samples with judgment results < 90% are used for odor identification. The system's own algorithm can be used or the relevant algorithm can be used for self-calculation.

[0082] Authenticity identification: The electronic nose odor recognition data of 40s, 50s, and 60s of agarwood detection is input into the clustering recognition model of authenticity samples. If all three algorithms are judged to be true, the test sample is judged to be authentic agarwood. If one or more algorithms are judged to be false, the test sample is judged to be counterfeit agarwood.

[0083] Preferably, the odor data parameters include: 40s, 50s, 60s of electronic nose recognition data.

[0084] The beneficial technical effects of the present invention compared with the prior art are as follows:

[0085] The method of the present invention does not require any destructive actions such as cutting, grinding, crushing, drilling, and extraction of the agarwood sample, thereby avoiding damage to the authentic expensive agarwood.

[0086] The present invention adopts a technology that combines microscopic imaging and odor analysis. Through the microscopic system, it can achieve accurate imaging of agarwood oil and wood, and can have a strong ability to distinguish counterfeit and inferior agarwood made by common counterfeiting methods such as non-agarwood, high polishing, splicing, and high pressure; through the electronic nose odor analysis method, it can identify counterfeit and inferior agarwood made by counterfeiting methods such as micro-pressure, oil injection, medicine soaking, and agarwood paste smearing.

[0087] The present invention optimizes the magnification and imaging area of ​​microscopic imaging to ensure that the imaging clarity meets the identification requirements; and determines the analysis algorithm of electronic nose data to meet the requirements for authenticity identification of agarwood.

[0088] The rapid and non-destructive agarwood authenticity identification method proposed by the present invention can realize the authenticity identification of agarwood beads, agarwood pendants and small-volume agarwood handicrafts of any shape and size, and the process is non-destructive, rapid and highly accurate.

[0089] The existing imaging system comprises a lower rod, the upper end of which is provided with a first-axis gimbal, the first-axis gimbal is connected to a second-axis gimbal, and an imaging body is rotatably arranged in the second-axis gimbal. Figure 7 As shown, in the prior art, the first-axis gimbal and the second-axis gimbal cannot move toward or away from each other in the vertical direction, so the distance between the imaging body and the collected object cannot be adjusted by the second-axis gimbal, so that the imaging body cannot be in the best collection position, resulting in inaccurate collection results.

[0090] In order to solve the above problems, in one implementation, Figure 8 and Fig. 9 As shown, the imaging system includes a lower rod 1, and a first-axis gimbal 2 connected to the upper end of the lower rod 1, the free end of the first-axis gimbal 2 is connected to the second-axis gimbal 3, the second-axis gimbal 3 is in a "U"-shaped structure, an imaging body 4 is rotatably arranged in the second-axis gimbal 3, a sensor 5 is arranged on the lower end surface of the imaging body 4, and also includes a height lifting mechanism 6 for connecting the first-axis gimbal 2 and the second-axis gimbal 3. When working, the sensor 5 collects the vertical distance signal between the lower end of the imaging body 4 and the collected object, and adjusts the position of the imaging body 4 in the vertical direction through the height lifting mechanism 6 according to the vertical distance signal to adjust the vertical distance between the lower end of the imaging body 4 and the collected object.

[0091] In one embodiment, Fig. 9As shown, the height lifting mechanism 6 includes a connecting component 61 connected to the upper end surface of the second-axis gimbal 3, a first lifting rod 62 is fixedly connected to the upper end surface of the connecting component 61, a supporting ring 63 is movably sleeved outside the first lifting rod 62, a first fixing plate 64 is fixedly connected to the outer wall of the supporting ring 63, a free end of the first fixing plate 64 is fixedly connected to the second fixing plate 65, the second fixing plate 65 is fixedly arranged on the upper end surface of the first-axis gimbal 2, a first fixing block 66 is fixedly connected to the upper end of the first lifting rod 62, a third fixing plate 67 is fixedly connected to the side wall of the first fixing block 66, a second fixing block 68 is fixedly connected to the free end of the third fixing plate 67, a second lifting rod 69 is fixedly connected to the lower end surface of the second fixing block 68, the lower end of the second lifting rod 69 is fixedly connected to the driving end of the driving cylinder 610, the driving cylinder 610 is fixedly arranged on the first-axis gimbal 2, wherein the first-axis gimbal 2 and the second-axis gimbal 3 are staggered in the vertical direction.

[0092] The arrangement of the height lifting mechanism 6 in this embodiment can adjust the distance between the imaging body 4 and the collected object by adjusting the relative positions of the first axis gimbal 2 and the second axis gimbal 3 in the vertical direction; and the sensor 5 is also arranged in this example (it should be noted that the sensor 5 includes a sensor body and a processor), and the arrangement of the sensor 5 can adjust the distance between the imaging body 4 and the collected object in real time. During operation, when the sensor 5 detects that the distance between the imaging body 4 and the collected object is not the standard distance, the sensor 5 will start the driving cylinder 610, and the driving cylinder 610 will drive the second axis gimbal 3 to move upward or downward in the vertical direction relative to the first axis gimbal 2 through the combined structure of the second lifting rod 69, the second fixed block 68, the third fixed plate 67, the first fixed block 66, the first lifting rod 62 and the connecting assembly 61, and then the imaging body 4 is moved upward or downward through the second axis gimbal 3, thereby changing the distance between the imaging body 4 and the collected object so that the distance reaches the standard distance.

[0093] In order to enable the second-axis gimbal 3 to perform vertical movement relative to the first-axis gimbal 2 and also have a rotation function, in one embodiment, as Fig.10 As shown, the connecting component 61 includes a mounting base 613 fixedly connected to the second-axis gimbal 3, the mounting base 613 is slidably connected to the first-axis gimbal 2, and a first rotating rod 612 is fixedly connected to the upper end surface of the mounting base 613. The upper end of the first rotating rod 612 is fixedly connected to the driving end of the driving motor 611, and the driving motor 611 is fixedly arranged on the lower end of the first lifting rod 62.

[0094] The setting of the connecting component 61 in this embodiment can cause the second-axis gimbal 3 to rotate relative to the first-axis gimbal 2. Specifically, when it is necessary to collect different azimuthal features of the collected object on the horizontal plane, the sensor 5 will start the driving motor 611. The driving motor 611 will rotate the mounting base 613 through the first rotating rod 612, thereby rotating the second-axis gimbal 3 relative to the first-axis gimbal 2 through the mounting base 613, and then rotating the imaging body 4 relative to the collected object through the second-axis gimbal 3.

[0095] In one embodiment, Fig.11 As shown, the imaging system also includes a position locking mechanism 7 for mutually locking the mounting base 613 and the first-axis gimbal 2, the position locking mechanism 7 includes a plurality of position locking grooves 71 provided on the mounting base 613, a position locking rod 72 is provided in one of the position locking grooves 71, the position locking rod 72 is movably provided in a first movable groove 73, the first movable groove 73 is provided on the first-axis gimbal 2, one end of the position locking rod 72 located in the first movable groove 73 is fixedly connected to a first magnet block 74, the first magnet block 74 is movably inserted in the first movable groove 73, and a first electromagnet 76 matching the first magnet block 74 is fixedly connected to the side wall of the first movable groove 73.

[0096] The position locking mechanism 7 in this embodiment is arranged such that when the distance and angle between the imaging body 4 and the collected object are adjusted to the standard position, the current position of the imaging body 4 is locked by locking the mounting base 613 and the first-axis gimbal 2. It should be noted that no matter when the first-axis gimbal 2 and the second-axis gimbal 3 are adjusted in the vertical direction, or when the first-axis gimbal 2 and the second-axis gimbal 3 are adjusted for relative rotational motion, the position locking mechanism 7 does not lock the mounting base 613 and the first-axis gimbal 2. During operation, when it is necessary to adjust the first-axis gimbal 2 and the second-axis gimbal 3 in the vertical direction, or when the first-axis gimbal 2 and the second-axis gimbal 3 are adjusted for relative rotational motion, the first electromagnet 76 will be energized to generate magnetism, and under the action of the magnetic attraction, the first magnet block 74 will move the position locking rod 72 to the right, so that the position locking rod 72 moves out of the position locking slot 71, thereby realizing the unlocking process.

[0097] It should be noted that several position locking grooves 71 are arranged in a ring shape, and there is a gap between two adjacent position locking grooves 71, and the size of the gap is less than 0.5 mm. This arrangement can enable the position locking rod 72 to accurately enter the position locking groove 71; in order to enable the position locking rod 72 to more accurately enter the position locking groove 71, in one example, the left end of the position locking rod 72 is set to a round head structure, and in another example, the left end of the position locking rod 72 is set to a cone structure; after the position locking rod 72 enters the position locking groove 71, in order to enable the position locking groove 71 to firmly contact the position locking rod 72, the groove cavity of the position locking groove 71 is set to a cone structure.

[0098] In one embodiment, Fig.11 As shown, the position locking mechanism 7 also includes a first guide support assembly 75 which is arranged between the first magnet block 74 and the first electromagnet 76 and is used to connect the first magnet block 74 and the first electromagnet 76. The first guide support assembly 75 includes a first guide support rod 751, a first guide support tube 752 and a first guide support spring 753. The first guide support rod 751 is movably inserted in the first guide support tube 752, and the first guide support spring 753 is wound around the outside of the first guide support rod 751, and the two ends of the first guide support spring 753 are respectively fixedly connected to the side wall of the first guide support rod 751 and the outer side wall of the first guide support tube 752.

[0099] The provision of the first guide support assembly 75 in this embodiment can both serve as a guide and support for the movement of the position locking rod 72 so that the position locking rod 72 only moves in the horizontal direction, and at the same time provide a reset force for the reset movement of the position locking rod 72.

[0100] When the first-axis gimbal 2 and the second-axis gimbal 3 are adjusted in the vertical direction, or when the first-axis gimbal 2 and the second-axis gimbal 3 are adjusted for relative rotation, in order to prevent the imaging body 4 from moving relative to the second-axis gimbal 3, in one embodiment, Fig.12As shown, the imaging system also includes a fastening mechanism 8, which includes a rigid fastening plate 81 and an elastic fastening plate 82 connected to each other, the elastic fastening plate 82 is arranged close to the imaging body 4, the rigid fastening plate 81 and the elastic fastening plate 82 are movably arranged in the second moving groove 83, the second moving groove 83 is opened on the second axis gimbal 3, a first moving rod 84 is fixedly connected to the side wall of the rigid fastening plate 81, the first moving rod 84 is movably plugged into the second axis gimbal 3, the free end of the first moving rod 84 is fixedly connected to the second magnet block 85, the side wall of the second axis gimbal 3 is fixedly connected to the second electromagnet 86 matching the second magnet block 85, and the second electromagnet 86 is fixedly connected to the side wall of the second axis gimbal 3. The magnet 86 is movably mounted outside the first moving rod 84, and the side wall of the second magnet block 85 is fixedly connected to the second guide support assembly 87. The free end of the second guide support assembly 87 is fixedly connected to the second-axis gimbal 3. The second guide support assembly 87 includes a second guide support rod 871, a second guide support tube 872 and a second guide support spring 873. The second guide support rod 871 is movably inserted in the second guide support tube 872, and the second guide support spring 873 is wound and connected to the outside of the second guide support rod 871, and the two ends of the second guide support spring 873 are respectively fixedly connected to the side wall of the second guide support rod 871 and the outer side wall of the second guide support tube 872.

[0101] The setting of the fastening mechanism 8 in this embodiment can prevent the imaging body 4 and the second axis gimbal 3 from moving relative to each other when the first axis gimbal 2 and the second axis gimbal 3 are adjusted in the vertical direction, or when the first axis gimbal 2 and the second axis gimbal 3 are adjusted in relative rotational motion, through the fastening effect. During operation, when the first axis gimbal 2 and the second axis gimbal 3 are adjusted in the vertical direction, or when the first axis gimbal 2 and the second axis gimbal 3 are adjusted in relative rotational motion, the second electromagnet 86 is energized to generate magnetism, and under the action of magnetic attraction, the second magnet block 85, the first moving rod 84 and the rigid fastening plate 81 are pressed against the outer wall of the imaging body 4 with the elastic fastening plate 82, thereby fixing the imaging body 4 and the second axis gimbal 3 together through the interference effect.

[0102] In one embodiment, Fig.12As shown, the imaging system also includes a heat dissipation mechanism 9, which includes a connecting plate 91 connected to the second magnet block 85, and a second moving rod 92 is fixedly connected to the side wall of the connecting plate 91. A driving screw rod 93 is fixedly sleeved outside the second moving rod 92, and the driving screw rod 93 is movably inserted into the second-axis gimbal 3. The driving screw rod 93 is externally threaded with an internally threaded tube 94, and the internally threaded tube 94 is rotatably set on the side wall of the second-axis gimbal 3. A first driving gear 95 is fixedly sleeved outside the internally threaded tube 94, and the first driving gear 95 and the second driving gear 96 are meshed with each other. The second driving gear 96 is fixedly sleeved outside the second rotating rod 97, and one end of the second rotating rod 97 is rotatably set on the side wall of the second-axis gimbal 3, and the other end of the second rotating rod 97 is fixedly connected to a rotating fan 98.

[0103] The heat dissipation mechanism 9 in this embodiment can dissipate heat for the device. When the first-axis gimbal 2 and the second-axis gimbal 3 are adjusted in the vertical direction, or when the first-axis gimbal 2 and the second-axis gimbal 3 are adjusted in relative rotation, the device will generate heat. The heat dissipation mechanism 9 can achieve a high-efficiency heat dissipation effect. Specifically, the rightward movement of the second magnet block 85 will drive the driving screw rod 93 to move rightward through the connecting plate 91 and the second moving rod 92. The movement of the driving screw rod 93 will cause the internal threaded tube 94 to rotate. The rotation of the internal threaded tube 94 will cause the rotating fan 98 to rotate through the first driving gear 95, the second driving gear 96 and the second rotating rod 97. The rotation of the rotating fan 98 will accelerate the flow of natural wind, thereby achieving a heat dissipation effect on the device.

[0104] It should be noted that in this embodiment, the wheel diameter of the first driving gear 95 is much larger than the wheel diameter of the second driving gear 96, so as to ensure that when the first driving gear 95 rotates one circle, the second driving gear 96 can rotate multiple circles.

[0105] Although the embodiments of the present invention are described in detail above, it is obvious to those skilled in the art that various modifications and variations can be made to these embodiments. However, it should be understood that such modifications and variations are within the scope and spirit of the present invention as described in the claims. Moreover, the present invention described herein may have other embodiments and may be implemented or realized in a variety of ways.

Claims

1. A method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose, characterized in that: The method comprises the following steps: S1: Use microscope and imaging system to obtain microstructure images of agarwood samples; S2: Input the microstructure image of the agarwood test sample into the agarwood authenticity identification model based on microscopic images. If the judgment result is ≥90%, it is authentic agarwood, and the test is stopped. If the judgment result is <90%, the authenticity identification based on odor analysis is performed; S3: According to the agarwood authenticity identification model based on smell, the data of the agarwood test sample in each sensor of the electronic nose is obtained; S4: Turn on the automated pattern recognition system of the electronic nose, input the electronic nose odor recognition data of 40s, 50s, and 60s of agarwood detection into the clustering recognition model of the authenticity sample, and use the Euclidean algorithm, the deterministic finite automatic algorithm, and the segmented expansion algorithm to give the discrimination results. If all three algorithms are judged to be true, the detection sample is judged to be authentic agarwood. If one or more algorithms are judged to be false, the detection sample is judged to be counterfeit agarwood.

2. The method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose according to claim 1 is characterized in that: The agarwood authenticity identification model based on microscopic images comprises the following steps: S11: Use microscope and imaging system to obtain microstructure images of authentic agarwood samples; S12: extracting features from the image data, extracting features from the sample microstructure image, and obtaining features such as pores, internal phloem, wood fibers, or wood rays of the sample, as well as features such as surface finish, color, oil distribution, and oil overflow; S13: Obtain the microstructure image feature vector of the authentic and fake agarwood samples; train the sample feature vector using a set algorithm to obtain an agarwood recognition model.

3. The method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose according to claim 1 is characterized in that: The agarwood authenticity identification model based on smell comprises the following steps: S31: Place the agarwood sample in a sealed container and leave it at room temperature for 30 to 60 minutes, or shake the sample at 60°C for 10 to 20 minutes; S32: Turn on the electronic nose, preheat the device for 20 to 30 minutes, insert the injection needle of the sensor into the soft cover of the sealed injection bottle, the distance between the injection needle and the bottle mouth is 1 / 3 of the bottle height, and then insert the air inlet needle of the sensor into the soft cover of the injection bottle, the distance between the injection needle and the bottle mouth is greater than the distance between the injection needle and the bottle mouth by 1mm; then set the array parameters of the sensor for detection: gas flow rate 0.5-0.7L / min, cleaning time 50-100s, detection time 80-100s, start sampling and measurement at 15-30℃, and obtain the data of the authenticity detection sample in each sensor of the electronic nose respectively; S33: Turn on the automatic pattern recognition system of the electronic nose, and use linear discriminant analysis to cluster the electronic nose gas recognition data of 40s, 50s, and 60s of the authentic and fake agarwood samples, respectively, to obtain a cluster recognition model for the authentic agarwood samples.

4. The method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose according to claim 1 is characterized in that: The imaging system comprises a lower rod (1), and a first-axis pan-tilt platform (2) connected to the upper end of the lower rod (1); the free end of the first-axis pan-tilt platform (2) is connected to a second-axis pan-tilt platform (3); the second-axis pan-tilt platform (3) is in a "U"-shaped structure; an imaging body (4) is rotatably arranged inside the second-axis pan-tilt platform (3); a sensor (5) is arranged on the lower end surface of the imaging body (4); and a height lifting mechanism (6) for connecting the first-axis pan-tilt platform (2) and the second-axis pan-tilt platform (3); when in operation, the sensor (5) collects a vertical distance signal between the lower end of the imaging body (4) and the collected object, and adjusts the position of the imaging body (4) in the vertical direction through the height lifting mechanism (6) according to the vertical distance signal, so as to adjust the vertical distance between the lower end of the imaging body (4) and the collected object.

5. The method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose according to claim 4 is characterized in that: The height lifting mechanism (6) includes a connecting component (61) connected to the upper end surface of the second-axis gimbal (3), a first lifting rod (62) is fixedly connected to the upper end surface of the connecting component (61), a supporting ring (63) is movably sleeved outside the first lifting rod (62), a first fixing plate (64) is fixedly connected to the outer side wall of the supporting ring (63), a free end of the first fixing plate (64) is fixedly connected to a second fixing plate (65), the second fixing plate (65) is fixedly arranged on the upper end surface of the first-axis gimbal (2), and the upper end of the first lifting rod (62) is fixedly connected to the second fixing plate (65). The end is fixedly connected to a first fixed block (66), the side wall of the first fixed block (66) is fixedly connected to a third fixed plate (67), the free end of the third fixed plate (67) is fixedly connected to a second fixed block (68), the lower end surface of the second fixed block (68) is fixedly connected to a second lifting rod (69), the lower end of the second lifting rod (69) is fixedly connected to the driving end of a driving cylinder (610), and the driving cylinder (610) is fixedly set on the first-axis pan-tilt platform (2), wherein the first-axis pan-tilt platform (2) and the second-axis pan-tilt platform (3) are staggered in the vertical direction.

6. The method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose according to claim 5 is characterized in that: The connecting component (61) includes a mounting base (613) fixedly connected to the second-axis pan-tilt head (3), the mounting base (613) is slidably connected to the first-axis pan-tilt head (2), and a first rotating rod (612) is fixedly connected to the upper end surface of the mounting base (613), and the upper end of the first rotating rod (612) is fixedly connected to the driving end of the driving motor (611), and the driving motor (611) is fixedly arranged on the lower end of the first lifting rod (62).

7. The method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose according to claim 6 is characterized in that: The imaging system further comprises a position locking mechanism (7) for mutually locking the mounting base (613) and the first-axis gimbal (2), the position locking mechanism (7) comprising a plurality of position locking grooves (71) provided on the mounting base (613), wherein a position locking rod (72) is provided in abutting relation within one of the position locking grooves (71), the position locking rod (72) being movably provided within a first movable groove (73), the first movable groove (73) being provided on the first-axis gimbal (2), one end of the position locking rod (72) located within the first movable groove (73) being fixedly connected to a first magnet block (74), the first magnet block (74) being movably inserted into the first movable groove (73), and a first electromagnet (76) matching the first magnet block (74) being fixedly connected to a side wall of the first movable groove (73).

8. The method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose according to claim 7 is characterized in that: The position locking mechanism (7) also includes a first guide support assembly (75) which is arranged between the first magnet block (74) and the first electromagnet (76) and is used to connect the first magnet block (74) and the first electromagnet (76). The first guide support assembly (75) includes a first guide support rod (751), a first guide support tube (752) and a first guide support spring (753). The first guide support rod (751) is movably inserted into the first guide support tube (752), and the first guide support spring (753) is wound around the outside of the first guide support rod (751), and the two ends of the first guide support spring (753) are respectively fixedly connected to the side wall of the first guide support rod (751) and the outer side wall of the first guide support tube (752).

9. The method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose according to claim 8, characterized in that: The imaging system further comprises a fastening mechanism (8), wherein the fastening mechanism (8) comprises a rigid fastening plate (81) and an elastic fastening plate (82) connected to each other, wherein the elastic fastening plate (82) is arranged close to the imaging body (4), wherein the rigid fastening plate (81) and the elastic fastening plate (82) are movably arranged in a second movable groove (83), wherein the second movable groove (83) is provided on the second axis gimbal (3), wherein a first movable rod (84) is fixedly connected to the side wall of the rigid fastening plate (81), wherein the first movable rod (84) is movably plugged into the second axis gimbal (3), wherein a second magnet block (85) is fixedly connected to the free end of the first movable rod (84), wherein a second electromagnet (86) matching the second magnet block (85) is fixedly connected to the side wall of the second axis gimbal (3), and wherein the second electromagnet (86) matching the second magnet block (85) is fixedly connected to the side wall of the second axis gimbal (3). The electromagnet (86) is movably mounted outside the first moving rod (84); the side wall of the second magnet block (85) is fixedly connected to a second guide support assembly (87); the free end of the second guide support assembly (87) is fixedly connected to the second axis pan / tilt platform (3); the second guide support assembly (87) comprises a second guide support rod (871), a second guide support tube (872) and a second guide support spring (873); the second guide support rod (871) is movably inserted into the second guide support tube (872); the second guide support spring (873) is wound around and connected to the outside of the second guide support rod (871); and the two ends of the second guide support spring (873) are respectively fixedly connected to the side wall of the second guide support rod (871) and the outer side wall of the second guide support tube (872).

10. The method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and electronic nose according to claim 9, characterized in that: The imaging system further comprises a heat dissipation mechanism (9), the heat dissipation mechanism (9) comprising a connecting plate (91) connected to the second magnet block (85), the side wall of the connecting plate (91) being fixedly connected to a second moving rod (92), the second moving rod (92) being fixedly sleeved with a driving screw rod (93) outside, the driving screw rod (93) being movably plugged into the second axis gimbal (3), the driving screw rod (93) being externally threadedly sleeved with an internally threaded tube (94), the internally threaded tube (94) being rotatably arranged on the side wall of the second axis gimbal (3), the internally threaded tube (94) being externally fixedly sleeved with a first driving gear (95), the first driving gear (95) and the second driving gear (96) being meshed with each other, the second driving gear (96) being fixedly sleeved outside a second rotating rod (97), one end of the second rotating rod (97) being rotatably arranged on the side wall of the second axis gimbal (3), the other end of the second rotating rod (97) being fixedly connected to a rotating fan (98).

Images