CN106228167B - Collection apparatus, mark and recognition methods - Google Patents

Abstract

The invention discloses a feature acquisition, marking and identification method. The acquisition method includes using three-dimensional scanning to reconstruct a high-precision three-dimensional model of an object to be measured to obtain a complete and precise geometric model of the target surface; collecting high-resolution photos of the surface of the object to be measured; calculating The mapping relationship between high-resolution photos and precise geometric models, mapping high-resolution photos to precise geometric models to obtain high-precision surface color models; extracting the three-dimensional geometric features of high-precision surface color models; Acquisition of surface constant optical features; acquisition of acoustic features of the object to be tested; use of three-dimensional geometric features, surface constant optical features, and acoustic features or fusion of two or more to express the unique feature set of the object to be measured; record the feature set to in the database. The present invention is applied to feature collection, marking and identification of valuables, such as antiques and works of art, so as to improve their collection value and eliminate the possibility of imitation.

Description

Feature Acquisition, Labeling and Recognition Methods

technical field

The present invention relates to a feature acquisition, marking and identification method, specifically based on the collection and analysis technology of full-dimensional fusion information to collect, mark and identify the unique features of the object to be tested, that is, all substances with entities.

Background technique

China is a country with a large cultural heritage and has many antiques and artworks, such as precious cultural relics, calligraphy and paintings, handmade artworks by masters, jewelry and diamonds, etc. Correspondingly, many collectors have emerged, which has given birth to the transaction of antiques and artworks market. In the trading market, authenticity and originality are undoubtedly the key factors affecting the transaction price and collection significance.

At present, the common identification methods on the market include two categories: physical identification and scientific and technological identification. Physical identification is a means of visual inspection by employing well-known experts in the industry, and using manual experience to determine the shape, color and other details of antiques and artworks. Experts are required to be senior enough and have rich identification experience, which is very subjective. Strong identification method. Scientific and technological appraisal refers to the use of analytical instruments to detect the chemical composition of antiques and artworks, and to conduct comparative analysis of antiques and artworks based on the existing database of antiques and artworks of different ages. Common analysis methods include energy dispersive X-ray fluorescence analysis (EDX) and proton excited X-ray fluorescence analysis (pixe). The detection process is to put antiques and artworks or collected samples into the instrument, select the appropriate spectral head according to the targets of different textures, and after vacuuming, the machine starts to detect fluorescence at 3-5 mm for spectral testing. Multiple tests are carried out on the surface, so as to measure the chemical composition or substance content contained in the target, and finally compare the detected data with the database of the authoritative department, and give the identification conclusion.

The advantage of scientific and technological identification is that it can overcome subjective errors, and has good accuracy and reproducibility, but it requires material sampling of the target and large-scale database support. In well-known appraisal institutions, a combination of physical appraisal and technological appraisal is generally used to give multiple appraisal opinions with sufficient persuasiveness and scientific basis. Whether it is physical identification or scientific and technological identification, after the identification, collectors or traders get conclusion reports on authenticity or certificates on antiques and works of art, but these reports and certificates have the possibility of fraud. Even after a dispute arises, it is difficult to determine whether it is the original.

The above-mentioned physical identification and scientific and technological identification have the following disadvantages. The identification procedure is complicated, and various equipment needs to be used to identify the object to be tested by different procedures, and sometimes it will cause certain damage to the object to be tested; In other words, the appraisal fee is high and unacceptable; even if the appraisal is passed, it is difficult for both the buyer and the seller to determine "this thing is the other thing", and the received item is not an appraisal item; physical appraisal requires the appraiser to have rich professional knowledge and According to experience, the cost of identification equipment required for scientific and technological identification is high, and the identification cost is too high.

Therefore, if the appearance, shape, and structure of antiques and artworks can be reconstructed in multiple dimensions after the identification is completed, that is, unique identification can be performed through shape, sound, and light waves. Realize the preservation of digitization, design a rigorous method to extract the unique characteristics of the antiques and artworks, that is, only the original has this unique characteristic, even if the counterfeit and imitation are very similar, they can pass this unique characteristic. sexual "identity" characteristics. The record of this unique feature can undoubtedly improve the collection value of high-end valuables and reduce the possibility of imitation.

This shows that above-mentioned relevant identification method obviously still has inconvenience and defect in use, and needs urgently to be further improved. In order to solve the above-mentioned problems, the relevant manufacturers have tried their best to find a solution, but for a long time no suitable design has been developed, and the general product has no suitable structure to solve the above-mentioned problems. This problem is obviously related The problem that the industry is eager to solve.

Contents of the invention

The technical problem to be solved by the present invention is to provide a feature collection, marking and identification method, which can obtain the uniqueness of the object to be measured by collecting one or more of the three-dimensional geometric features, surface constant optical features, and acoustic features of the object to be measured. A feature set that increases the collection value of the object under test such as antiques, artwork and all valuables and reduces the possibility of imitation.

The purpose of the present invention and the solution to its technical problems are achieved by adopting the following technical solutions.

The invention discloses a feature acquisition method, which includes the following steps: step S1, using three-dimensional scanning to reconstruct a high-precision three-dimensional model of the object to be measured, to obtain a complete and accurate geometric model of the surface of the object to be measured; step S2, collecting the surface of the object to be measured High-resolution photos; step S3, calculate the mapping relationship between the high-resolution photos and the precise shape geometry model, map the high-resolution photos to the precise shape geometry model, and obtain a high-precision surface color model; step S4, extract high-precision The three-dimensional geometric features of the surface color model; step S5, in a constant lighting environment, the surface constant optical features of the object to be measured are collected; step S6, using one of the three-dimensional geometric features and surface constant optical features or combining them , to obtain a feature set expressing the uniqueness of the analyte; step S7, to record the feature set into the database.

The purpose of the present invention and its technical problems can also be further realized by adopting the following technical measures.

In the aforementioned collection method, step S5 includes: 1) irradiating the object to be measured with the irradiating light and recording the data related to the current lighting environment; 2) collecting the reflection values of the surface of the object to be measured to the irradiating light in different directions or/and at different positions; 3 ) after performing digital sampling, record the quantitative characteristics as the unique surface constant optical characteristics of the object to be tested; and the reflection value of the irradiated light includes one or both of reflection value, penetration value, absorption value, diffraction value, etc. more than one combination.

The aforementioned acquisition method, after step S5 and before step S6, also includes step S51: 1) transmit an acoustic signal to the object to be measured and record the relevant data of the current acoustic environment, when the acoustic signal passes through the object to be measured, stop emitting the acoustic signal; 2) collect Measure the reflection value of the object under test to the acoustic signal in the environment; 3) after digital sampling, record the quantitative feature as the unique acoustic feature of the object under test; in step S6, use the acoustic feature, the three-dimensional geometric feature , one of the constant optical features of the surface or a combination of two or more to obtain a feature set that expresses the uniqueness of the object to be measured; and wherein the reflected value of the acoustic signal includes a reflection value, a penetration value, an absorption value, and a diffraction value. species or a combination of two or more.

In the aforementioned acquisition method, step S1 includes obtaining a point cloud model expressing the shape of the object to be measured through the three-dimensional scanning, and the error of the point cloud is less than 0.01mm.

In the aforementioned acquisition method, the three-dimensional geometric features in step S4 include points, line segments, curves, polygonal planes and curved surfaces.

In the aforementioned acquisition method, the three-dimensional geometric features described in step S4 include self-defined three-dimensional geometric features based on the characteristics of the object to be measured.

In the aforementioned acquisition method, the extraction of points in step S4 includes: first determining the image point position of the point from the high-resolution photo with sub-pixel extraction technology, and then mapping the image point position to the precise geometric model, and finally determining the position of the image point on the precise geometric model. The exact position is obtained by extracting neighborhood points from the model, and the attributes of point features include spatial XYZ coordinates; the extraction of line segments in step S4 includes: determining two endpoints with sub-pixel extraction technology from a single or multiple high-resolution photos The position of the image point, and then map the position of the image point to the precise shape geometry model, then extract the neighborhood points on the precise shape geometry model to fit the exact position of the end point, and finally connect the two points on the precise shape geometry model To form a line segment, the attributes of the line segment feature include the spatial coordinates and the length of the two endpoints; and the extraction of the curve in step S4 includes: determining multiple samples located on the curve with sub-pixel extraction technology from a single or multiple high-resolution photos The image point position of the point, and then map the image point position to the precise shape geometric model, and then extract the neighborhood point fitting on the precise shape geometry model to obtain the exact position of the sampling point, and finally multiple sampling The points are re-fitted according to the mathematical model, and the attributes of the curve feature include curve length and curve parameters.

In the aforementioned acquisition method, the unique feature set in step S6 is composed of encrypted unique feature codes.

The purpose of the present invention and the solution to its technical problems can also be achieved by adopting the following technical solutions.

The invention discloses a feature marking method, which adopts the aforementioned collection method to collect data, and archives and records the collected data together with the historical information of the object to be tested.

The purpose of the present invention and the solution to its technical problems can also be achieved by adopting the following technical solutions.

The present invention discloses a feature recognition method comprising the following steps: step S11, using the aforementioned acquisition method, performing unique feature encoding calculation on the review object according to the same acquisition rules in the same measurement environment, and obtaining the unique feature of the review object set; step S12, compare the unique feature code in the unique feature set of the review object with the unique feature code of the corresponding original object to be tested stored in the database, and obtain the similarity ratio value; step S13, according to the similarity ratio The relationship between the value and the predetermined threshold determines whether the re-examination object is the original analyte of the original recorded data.

By means of the above technical solution, a feature acquisition, marking and identification method of the present invention has at least the following advantages and beneficial effects:

(1) The non-contact measurement method will not cause damage to the object to be measured;

(2) The full-dimensional fusion information can collect various signals such as optics, acoustics, and images to mark the object to be tested, and reflect the inherent characteristics of the object to the greatest extent;

(3) The extracted unique features can not only provide assistance for the re-identification of the test object, but also use the uniqueness to increase the value of its collection and identify imitations;

(4) The full-dimensional fusion information is automatically extracted by computer, without introducing manual subjective judgment, and reducing errors;

(5) The unique feature mark is encrypted, and can only be interpreted by using a special decryption algorithm, which increases the difficulty of cracking.

The above description is only an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention, it can be implemented according to the contents of the description, and in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable, The following preferred embodiments are described in detail as follows.

Detailed ways

In order to further explain the technical means and effects of the present invention to achieve the intended purpose of the invention, the specific implementation, structure, features and methods of a feature collection, marking and identification method proposed according to the present invention will be described below in conjunction with the preferred embodiments. Its effect is described in detail below.

In the present invention, the objects to be tested can be antiques, works of art, and all valuables and other objects with entities.

The invention provides a feature collection, marking and identification method of antiques and artworks based on full-dimensional fusion information. Full-dimension is a definition in the present invention that uses multiple technologies to collect signals in various ways from multiple sides and angles. The present invention specifically includes a method for collecting characteristics of the test object, a method for uniquely marking the test object, and a method for identifying the test object. Among them, the full-dimensional fusion information collection system platform is used for information collection of the object to be measured.

The full-dimensional fusion information collection system platform includes the following special equipment:

The physical feature acquisition equipment of the object to be tested: adopts active vision three-dimensional scanning to collect the physical features of the surface of the object to be tested, and obtains a precise point cloud model expressing the shape, and the error is less than 0.01mm to ensure that the physical features are sufficiently accurate.

High-definition image acquisition equipment: a high-definition image acquisition equipment composed of multiple high-resolution SLR cameras, which collects color and shape details of the target in different directions in the form of photos.

Acoustic signal transmitting equipment in a specific direction: transmit an acoustic signal of a certain frequency to the surface of the object to be measured in a specific direction, and the acoustic signal can propagate on the object to be measured, such as propagating on the surface and inner wall of the object to be measured and leaving a The unique acoustic signal reflects the value of the object to be measured.

Acoustic echo signal receiving equipment: collect the acoustic echo signal passing through the test object, such as the surface and inner wall of the test object, and perform acoustic digital sampling and recording of the signal strength and direction, that is, the reflection value of the test object to the acoustic signal.

Specific direction light emission equipment: emit light of a specific color and intensity to the target surface through light in a specific direction to create a constant lighting environment.

Surface optical reflection signal receiving equipment: collect the light intensity and signal reflected in different directions on the surface of the object to be measured, that is, the reflection value of the object to be measured to the irradiated light, and perform digital sampling and recording to obtain the optical characteristics of the target surface.

The information that can be obtained by using the full-dimensional fusion information collection platform is as follows:

Table 1

Embodiment 1, the feature collection method of the object to be measured

Next, the feature collection steps of the object to be tested in the present invention will be described in detail:

In step S1, the three-dimensional scanning is used to reconstruct the high-precision three-dimensional model of the object to be measured, so as to obtain a complete and accurate geometric model of the surface of the object to be measured.

The precise shape geometric model in step S1 is to use the surface shape feature acquisition equipment of the object to be measured to perform high-precision three-dimensional scanning of the object to be measured, and obtain a point cloud model that accurately expresses the shape of the object to be measured. The error of the point cloud is less than 0.01mm to ensure the acquisition of sufficiently accurate surface features.

Step S2, collecting high-resolution photos of the surface of the object to be measured.

Specifically, step S2 is to use the high-definition image acquisition device in the full-dimensional fusion information acquisition system to collect high-resolution photos of the surface of the object to be measured from multiple angles and all directions.

The high-resolution photo adopts any high-resolution that can be matched with the precise shape geometric model formed by the point cloud model with an error of less than 0.01mm, so that the high-resolution photo can fully reflect the surface shape characteristics of the object to be measured.

Step S3, calculating the mapping relationship between the high-resolution photo and the precise geometric model, mapping the high-resolution photo to the precise geometric model, and obtaining a high-precision surface color model. That is, multiple high-resolution photos are seamlessly connected without overlapping and one-to-one correspondence with precise geometric models.

Step S4, extracting the three-dimensional geometric features of the high-precision surface color model.

Specifically, in step S4, geometric features expressing shape information of key parts can be respectively extracted based on the precise shape geometry model. Common geometric features include points, line segments, curves, polygonal planes, and surfaces. In addition, for the special features on different objects to be tested, it is also possible to customize the three-dimensional geometric features based on the characteristics of the objects to be tested, such as antiques and artworks, and all valuables. Accurate geometric features can distinguish between authentic and fake test objects and (non-) original test objects based on the principle of equal physical features. Conventional geometric features are easy to imitate, but the unique geometric features defined by experts for each antique and artwork, and all valuables are often difficult to replicate, so it makes sense to use them for unique feature marks. The conventional and custom geometric features are respectively extracted from the precise shape geometry model and color model (that is, the color model formed by computer stored digital information) of the cultural relics in the collection, and these features are recorded in the database as a part of the full-dimensional fusion information. part. The following will describe the definition and extraction process of geometric features:

(1) Point features refer to single or multiple point objects with spatial XYZ coordinates. The extraction process is as follows: first determine the image point position of the point from the high-resolution photo with sub-pixel extraction technology, then map the image point position to the precise geometric model, and finally extract the neighborhood point fitting on the precise geometric model get the exact location. The attributes of a point feature include spatial XYZ coordinates.

(2) The line segment feature refers to the spatial connection of any two points on the target surface. The extraction process is as follows: from a single or multiple high-resolution photos, the image point positions of the two endpoints are determined by sub-pixel extraction technology, and then the image point positions are mapped to the precise shape geometric model, and then on the precise shape geometry model. The exact position of the end point is obtained by extracting neighborhood points and fitting, and finally the two points are connected on the precise shape geometry model to form a line segment. The attributes of the line feature include the spatial coordinates and length of the two endpoints.

(3) Curve features refer to features that can be expressed as second-order or higher-order curves on the target surface, such as circles, ellipses, and parabolas. The extraction process is as follows: from a single or multiple high-resolution photos, the image point positions of multiple sampling points located on the curve are determined by sub-pixel extraction technology, and then the image point positions are mapped to the precise geometric model, and then in the The exact position of the sampling points is obtained by extracting neighborhood points from the precise geometric model, and finally re-fitting multiple sampling points according to the mathematical model on the precise geometric model. Properties of regular curve features include curve length and curve parameters.

(4) Polygonal planar features refer to features that can be expressed as planes on the target surface, including regular rectangles, convex polygons, and concave polygons. The extraction process is as follows: from a single or multiple high-resolution photos, determine the image point positions of multiple endpoints on the edge of the polygon with sub-pixel extraction technology, and then map the image point positions to the precise geometric model, and then The exact position of the endpoint is obtained by extracting neighborhood points from the precise geometric model, and finally the endpoint is re-fitted into a plane on the precise geometric model. The attributes of the polygon plane feature include endpoint XYZ coordinates, area, side length.

(5) Surface features refer to features that can express second-order or higher-order surfaces on the target surface, such as spherical surfaces and ellipsoidal surfaces. The extraction process is as follows: from a single or multiple high-resolution photos, determine the image point positions of multiple sampling points on the edge of the surface with sub-pixel extraction technology, and then map the image point positions of the sampling points to the precise geometric model , and then extract the neighborhood point fitting on the precise shape geometry model to obtain the exact position of the endpoint, and then fit the model points surrounded by the endpoint into surface features. Properties of surface features include area, curvature, and surface parameters.

(6) Custom geometric features refer to the geometric shape features defined by experts based on experience to express the age or style of antiques and artworks, all valuables, such as the direction of strokes in calligraphy and painting artworks, and the engraved lines in porcelain artworks shape etc. The extraction process is as follows: from a single or multiple high-resolution photos, the sub-pixel extraction technology is used to determine the image point position of the sampling point on the custom shape, and then the image point position is mapped to the precise shape geometric model, and then in the precise The exact position of the sampling points is obtained by extracting neighborhood points from the geometric model, and finally connecting the sampling points on the precise geometric model to form a custom shape. The properties of custom geometric features include XYZ coordinates of sampling points, angle, line segment length, etc.

Step S5, under a constant illumination environment, collect surface constant optical features of the object to be tested.

Specifically, when collecting high-resolution photos of the object to be tested, the SLR camera can only obtain the color of the object to be tested under the current light source of the scene, not the inherent color of the object. But the human visual system has an important visual perception function - constant optical characteristics. This property ensures that humans perceive objects to maintain a relatively constant color under varying lighting conditions. The purpose of constant optical feature calculation is to eliminate the influence of different illumination on the true color of the surface of the object to be tested. A beam of light incident on the object to be tested is a probe. When light comes out of the DUT, it carries various information about the DUT. Conventional light measurements such as reflection, transmission, absorption, diffraction, etc., provide a lot of information about the object under test. However, the reflection value, penetration value, absorption value, and diffraction value of the same part of the object under the same lighting environment to the light irradiated in the same direction are constant and unique values that do not change over time. Partial detection data in a feature.

The light emitting device in a specific direction irradiates light of a specific color to the object to be tested. The light of a specific color is based on the constant optical characteristics of the human visual system, and is set according to the surface color, shape, and structure of the object to be tested. light of a specific color. The surface color of the object to be measured is made its original inherent color by the irradiation light of a specific color, and all the reflection values of the surface of the object to be measured to the irradiation light become the unique data of its identity. The light emitting device in a specific direction can, but is not limited to, irradiate light of a specific color to the object under test from different light-emitting points, different light intensities, different angles, and different positions.

The specific direction light emitting device in the full-dimensional fusion information collection system platform emits light of a specific color to the object to be measured to create a constant lighting environment, and collects the reflected signal strength in different directions through the surface optical reflection signal receiving device And direction, after digital sampling, record the quantitative characteristics as the constant optical characteristics of the surface of the object to be measured.

The constant lighting environment is a repeatable and reducible lighting environment, which makes the lighting environment of the re-examined object the same as that of the original object to be tested when the surface constant optical characteristics are collected again.

The constant lighting environment is hereinafter referred to as the lighting environment. The lighting environment includes but is not limited to various lighting methods artificially set by the tester according to different requirements and various lighting environments formed by various lighting equipment. The light emitting device in a specific direction can, but not limited to, irradiate the irradiating light to the object under test through one or a combination of two or more of the following aspects: 1) from which angles the irradiating light is emitted; 2) irradiating at each angle The intensity of light; 3) The angle of each light emitting point such as horizontal illumination, elevation angle illumination, top view angle illumination and other arbitrary angle illumination; 4) Increase the type and quantity of color filters to make the color of illumination light different; 5) The above-mentioned irradiation methods are implemented under the premise of different light settings; 6) The object to be tested is irradiated from different distances. The above only enumerates some irradiation methods, and also includes some other irradiation methods, which will not be repeated here. After the above settings are completed, the data is formed and stored in the database, and the measurement data of the object to be measured has uniqueness.

The data stored in the database for the above-mentioned lighting environment also includes but not limited to the temperature and humidity data of the environment and the object under test. If sweat enters the bracelet after wearing it for a certain period of time, it is necessary to take into account the corresponding moisture and the impact factors on the relevant tests after entering the bracelet.

Through the specific direction light emitting device in the full-dimensional fusion information collection system platform, the irradiation light is emitted to one or more specific positions of the object to be measured, which is used to create a lighting environment. At the same time, the value of the irradiated light, such as the intensity of light, the angle of light, the number of light-emitting points, side light, backlight and equipment relationship, etc. are recorded and stored in the database. The intensity and direction of the reflected light signal in different directions and/or different positions of the object to be tested are collected by the surface optical reflection signal receiving device, that is, the reflection value of the object to be measured to the irradiated light. Different types of objects to be tested have different response values to the irradiated light. The collected values of jade and glass objects include reflection value, penetration value, absorption value and diffraction value. The collection values of copperware, gold and silver jewelry to be measured are reflection value, diffraction value and absorption value. The reflected light and diffracted light of silk fabrics are very little, but the absorbed light is very large. The main collection of absorption values can also be combined with the collection of reflection values and diffraction values.

The reflection value of the irradiation light is a combination of one or more of the reflection value, penetration value, absorption value, and diffraction value. Other unlisted contrast irradiation lights that can reflect the uniqueness of the object to be measured can also be used. The reflected value of , will not be repeated here. After digital sampling, the quantitative characteristics are recorded as constant optical characteristics of the surface of the object under test.

For the same position of the object under test, the values obtained by the light irradiated from different directions are different. While collecting the above data, record the position information of the object under test and store it in the database. When collecting surface constant optical features, the object to be tested can be in any different positions, and the illumination light that irradiates the object to be measured can also be in any different positions. Multiple sets of unique characteristic data are obtained by combining different positions of the object to be measured and the irradiated light. It is also possible to set a unique lighting environment according to the specific conditions of the object to be measured to extract the unique identity value of the reflection value of the surface of the object to be measured to the irradiated light.

Step S51, transmitting an acoustic signal to the object to be measured, and collecting acoustic features of the object to be measured.

Any object has the function of sound transmission, such as glass, wood, steel, water, but its sound transmission effect is completely different. Different objects have different sound transmission effects. Just like a piano, they are all "strings" but their thickness, length, relaxation, etc. will change the sound. The sound is mainly divided into return, diffraction and refraction in the process of propagation. Through the three data, the unique identity of the object can be determined. This unique sound wave reflection is its unique identity.

Due to the different materials and internal structures of different objects to be tested, when an acoustic signal is sent to the object to be tested, the echo signals obtained are also different. In the same environment, such as the same temperature, pressure, and surrounding acoustic environment, when the same acoustic signal is applied to the exact same object under test at different times, the echo signals obtained, such as the reflection value of the object under test to the acoustic signal, The penetration value, diffraction value, absorption value, etc. are consistent. Therefore, the acoustic echo signal strength and direction of the same object under test measured by the same sound source in the same acoustic environment, that is, the reflection value of the object under test to the acoustic signal, is the reflection value, penetration value, diffraction value, absorption value, etc. The constant unique value does not change over time, and other unlisted values that reflect the uniqueness of the object to be measured can also be used to reflect the acoustic signal, and will not be repeated here.

The acoustic environment is created by the acoustic signal emitting equipment in a specific direction, that is, a dedicated test environment is set up. The data in this acoustic environment includes but is not limited to one or a combination of two or more of the following aspects. 1) The position and quantity of the sounding body and the determination of the sounding body; 2) The distance between the object to be measured and the sounding body; 3) The intensity of the sounding body; 4) Various sound values such as refraction, diffraction, absorption, sound wave, frequency, Sound quality, tone, pitch, sound intensity, etc.; 5) Acoustic environment and/or temperature and humidity of the object to be tested. Store the above data in the database. Transmitting an acoustic signal to the object to be tested, and storing part or all of the above-mentioned required information of the acoustic signal in a database.

When the acoustic signal passes through the object under test, stop emitting the acoustic signal. Turn on the acoustic echo signal receiving device to receive and collect the reflection value of the object under test to the acoustic signal in the measurement environment, that is, one or a combination of reflection value, penetration value, absorption value, and diffraction value. The test for collecting the penetration value is that the object under test and the acoustic echo signal receiving device are in the opposite direction of the same plane, that is, the 180-degree direction. There are various specific methods. You can use the sound method to test, or you can use the "standard sound bar" method to test, so I won't go into details here. The above data is digitally sampled and stored in the database, the quantitative characteristics are recorded, and the "unique identity data" of the object under test is formed as the unique acoustic feature of the object under test.

In the case of the same acoustic environment created by sending acoustic signals to the object under test when measuring acoustic characteristics, the values obtained by sending acoustic signals from different directions and different distances to the same position of the object under test are different. While collecting the above data, record the position information of the object under test and store it in the database. When collecting acoustic features, the object to be measured can be in any different positions, and the emitted acoustic signals can also be in any different positions and directions. Multiple sets of unique characteristic data are obtained by combining different positions and directions of the object under test and the emitted acoustic signal.

Step S6, adopting one or combining two or more of three-dimensional geometric features, surface constant optical features, and acoustic features to obtain a feature set expressing the uniqueness of the object to be measured.

Wherein, the unique feature set may be composed of encrypted unique feature codes. The unique feature encoding is to fuse one or more of the above-mentioned three-dimensional geometric features, surface constant optical features and acoustic features and encode them through an encryption algorithm to obtain a unique feature code, and add labels for convenience. Check later. This code can be restored to the above three characteristic signals after the decoding process.

In step S7, the feature set is recorded in the database, and the feature collection is completed.

Embodiment 2: Characteristic marking method of analytes

Next, the characteristic labeling method of the analyte of the present invention will be described in detail.

The present invention also proposes a feature marking method of the object to be tested, which uses the aforementioned feature collection method to collect unique data, and uses the historical information of the object to be tested as a mark feature to be archived and recorded together with the collected data. It plays an important role in proving whether antiques and works of art are orphans, and at the same time can increase the collection value of antiques and works of art.

Embodiment 3: Feature recognition method of the object to be tested

Next, the steps of the feature recognition method of the analyte of the present invention will be described in detail.

The present invention proposes a feature recognition method of an object to be measured, which includes the following steps:

Step S11 , using the above-mentioned feature collection method, to calculate the unique feature code of the object under test in the same measurement environment according to the same collection rules, and obtain the unique feature set of the re-examination object.

Step S12, comparing the unique feature codes in the unique feature set of the review object with the corresponding unique feature codes of the original object under test stored in the database to obtain a similarity ratio.

Identifying or detecting the analyte is also referred to as whether the analyte is the original analyte. Call out the corresponding detection data of the original analyte in the database by checking the label of the re-examination object. The detection data includes specific information data such as the specific environment, conditions, angles, and test points of the original object under test. According to these data, all reductive congruence tests are carried out on the re-examination object, that is, the re-examination object is tested under the same conditions as the original test object. Compare all the test data of the re-examination object with the test data of the original test object in the database to obtain a similarity ratio value less than or equal to 1. The larger the similarity, the closer to 1. Of course, other values can also be used. Different angles, test results are not the same. Same angle, different items, different results. Because it is a multi-angle test, even after the re-examination object is partially destroyed, it can still be judged by the data of other points.

Among them, the unique feature codes saved in the database are obtained in the following manner: After extracting the full-dimensional fusion information of various original objects to be tested according to the aforementioned embodiment one feature collection method, generate unique feature codes and record them in the database .

Step S13, according to the relationship between the similarity ratio and the predetermined threshold, determine whether the re-examined object is the original object under test.

No matter what method is used for the same object under test, it is difficult to achieve exactly the same sound. This is directly related to its degree of wear, sounding position, and relaxation. The purpose of the comparison is to avoid a misjudgment caused by an inconsistency. If several values in all the data are the same, even if one or two values are different but similar, it will not affect the overall judgment effect. For example, when testing with sound waves, it is necessary to restore the original state of the object under test, and compare the volume, pitch, sound intensity, timbre, and sound quality. These comparison values can ensure whether the test object can restore the original data.

Error exists in all tests, and strictly speaking no test method is error-free. The test process does not use manual analysis, but automatically compares all the re-test data after inputting them into the computer. The single-point test is to collect only one of the three-dimensional geometric features, surface constant optical features and acoustic features, or combine two or more of them to obtain a feature set that expresses the uniqueness of the review object. This feature set is compared with the original The unique feature set of the test object is compared. The multi-point test is to compare the above-mentioned characteristics collected at multiple locations of the re-examination object. For example, the predetermined threshold of the single-point test can be set to 99%, that is, the allowable error is 1%; the predetermined threshold of the multi-point test can be set to 95%, that is, the allowable error is 5%.

In step S12, a similarity ratio value less than or equal to 1 is defined. The larger the similarity, the closer to 1. Of course, it is not limited to this value, and other values can also be used.

For example, the similarity ratio value of the re-examination object in the single-point test is 0.98, and its percentage is 98% less than the predetermined threshold of 99%. A percentage of 98% is greater than the predetermined threshold of 95%, and the reexaminer is the original analyte.

The process used to assist the identification is as follows: Assume that there is a test object A and a test object B, both of which are consistent in appearance, and it is determined that the test object A is genuine, and now it is necessary to identify whether the test B is genuine or original. Analyte. The present invention does not identify the authenticity of the test object, but only tests and compares whether the original test object and the re-examination object are the same. After obtaining the unique feature code of the test object A by adopting the feature collection method of Embodiment 1, scan the test object B, calculate the unique feature code of the test object B, and compare the two codes. Within the range allowed by the difference threshold, a conclusion that the appearance of the test object B is consistent with the appearance of the test object A is given, otherwise a conclusion that the two are inconsistent is given. The conclusion is submitted to the user or the identification organization as one of the evidences for identifying whether two or more test objects are the same or whether the test objects are true or false.

Full-dimensional fusion information can be the fusion of 3D geometric features and surface constant optical features, or the fusion of 3D geometric features and acoustic features, or the fusion of surface constant optical features and acoustic features, or the fusion of 3D geometric features, acoustic One of the features and the surface constant optical features or the combination of the three, and the corresponding marking method and identification method can also be implemented in a corresponding manner.

The above description is only a preferred embodiment of the present invention, and does not limit the present invention in any form. Although the present invention has been disclosed as above with preferred embodiments, it is not intended to limit the present invention. Anyone familiar with this field Those skilled in the art, without departing from the scope of the technical solution of the present invention, can use the technical content disclosed above to make some changes or modify them into equivalent embodiments with equivalent changes, but any content that does not depart from the technical solution of the present invention, according to the present invention Any simple modifications, equivalent changes and modifications made to the above embodiments by the technical essence still belong to the scope of the technical solutions of the present invention.

Claims (8)

1. a kind of method for collecting characteristics, it is characterised in that include the following steps：

Step S1, carries out high-precision three-dimensional Model Reconstruction to determinand using 3-D scanning, it is completely smart to obtain determinand surface Quasi- body geometrical model；

Step S2, the high resolution picture on collection determinand surface；

Step S3, calculates the mapping relations between high resolution picture and accurate body geometrical model, high resolution picture is reflected Accurate body geometrical model is mapped to, obtains high-precision surface color model；

Step S4, extracts the three-dimensional geometry feature of high-precision surface color model；

Step S5, under invariable light environment, surface constant optical collection apparatus is carried out to determinand；Step S5 bags Include：1) irradiation light is irradiated to determinand and records current light context related data；2) different directions or/and different positions are gathered Put reflection numerical value of the determinand surface to irradiation light；3) after being digitized sampling, quantization characteristic is recorded, as determinand Uniqueness surface constant optical feature；And the reflection numerical value wherein to irradiation light includes reflected value, penetrating value, absorption value, spreads out Penetrate one kind or its two or more combination of value；

Step S51,1) launch acoustic signal to determinand and record current acoustic context related data, when acoustic signal is by treating After surveying thing, stop transmitting acoustic signal；2) reflection numerical value of the determinand to acoustic signal in measuring environment is gathered；3) numeral is carried out After changing sampling, quantization characteristic, the uniqueness acoustic feature as determinand are recorded；

Step S6, using three-dimensional geometry feature, one kind of surface constant optical feature or is merged, and obtains expression determinand only The feature set of one property, it is in step s 6, special using the acoustic feature, the three-dimensional geometry feature, the surface constant optical The feature set of one kind or fusion two of which expression determinand uniqueness achieved above of sign；And wherein to the anti-of acoustic signal Reflecting numerical value includes reflected value, penetrating value, absorption value, one kind of diffraction value or its two or more combination；

Step S7, feature set recorded in database.

2. acquisition method according to claim 1, it is characterised in that step S1 includes obtaining table by the 3-D scanning Up to the point cloud model of determinand shape, point cloud error is less than 0.01mm.

3. acquisition method according to claim 1, it is characterised in that three-dimensional geometry feature described in step S4 includes point, line Section, curve, polygon plane and curved surface.

4. acquisition method according to claim 1, it is characterised in that three-dimensional geometry feature described in step S4 includes being based on The self-defined three-dimensional geometry feature of determinand own characteristic.

5. acquisition method according to claim 3, it is characterised in that：

The extraction at step S4 midpoints includes：Image point position a little is first determined with sub-pixel detection technology from high resolution picture, Then image point position is mapped on accurate body geometrical model, the fitting of neighborhood point is finally extracted on accurate body geometrical model Accurate position is obtained, the attribute of point feature includes space X YZ coordinates；

The extraction of step S4 middle conductors includes：From individual or multiple high resolution pictures two are determined with sub-pixel detection technology Image point position, is then mapped on accurate body geometrical model by the image point position of endpoint, then in accurate body geometrical model Upper extraction neighborhood point is fitted to obtain the accurate position of endpoint, and two points finally are connected structure on accurate body geometrical model Into line segment, the attribute of line segment feature includes the space coordinate and length of two endpoints；

And the extraction of curve includes in step S4：Determined from individual or multiple high resolution pictures with sub-pixel detection technology Image point position, is then mapped on accurate body geometrical model, then by the image point position of multiple sampled points on curve Extract neighborhood point on accurate body geometrical model to be fitted to obtain the accurate position of sampled point, finally in accurate body geometrical model On multiple sampled points are fitted according to mathematical model again, the attribute of curvilinear characteristic includes length of curve and parameter of curve.

6. acquisition method according to claim 1, it is characterised in that the feature set of the uniqueness in step S6 is by encrypted Uniqueness characteristic coding is formed.

A kind of 7. signature method, it is characterised in that data acquisition is carried out using the acquisition method described in claim 1-6, and Record is filed together with the determinand historical information of itself to the data of collection.

8. a kind of characteristic recognition method, it is characterised in that include the following steps：

Step S11, using the acquisition method described in claim 1-6, to review thing according to identical under identical measuring environment Collection rule carry out uniqueness characteristic coding calculating, obtain review thing uniqueness characteristic collection；

Step S12, the uniqueness characteristic coding that the uniqueness characteristic of the review thing is concentrated are corresponding with what is preserved in database The uniqueness characteristic coding of former determinand is compared, and obtains similar rate value；

Step S13, according to the relation of similar rate value and predetermined threshold, determine the review thing whether be original recorded data original Determinand.