Automatic pairing system and method for tree seed amusement, and electronic equipment

Abstract

The invention provides an automatic pairing system and method for tree seed amusement items and electronic equipment, and the system comprises a collection device, an information processing device and a collection positioning storage cabinet, the collection device is used for collecting feature images of to-be-paired tree seed amusement items; the information processing device is used for determining feature information of the to-be-paired tree seed amusement items in different dimensions based on the feature images; the placing positions of the to-be-paired tree seed amusement items in the collecting and positioning storage cabinet are generated; based on the feature information, the placement positions and the pairing model, determining a target placement position of each paired tree seed amusement item in the paired tree seed amusement item set in the collection positioning storage cabinet and an evaluation index of at least one dimension corresponding to the paired tree seed amusement item set; and the collecting, positioning and storing cabinet is used for placing the to-be-paired tree seed amusement items according to the placing positions. According to the automatic pairing system for the tree seed amusement amusement, the pairing accuracy and pairing efficiency of the tree seed amusement amusement are improved, and the tree seed amusement amusement is quickly searched by using the collecting and positioning storage cabinet.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence pairing of tree seed-type collectibles, and in particular to an automatic pairing system, method and electronic equipment for tree seed-type collectibles. Background Art

[0002] Up to now, the raw seeds of tree seed-based ornamental items (including but not limited to hemp walnuts, iron walnuts, elm walnuts, lantern walnuts, etc.) represented by ornamental walnuts, need to be screened and matched by professional pairing masters after they are harvested from the trees. Finally, paired products or classified bracelets and handheld materials that meet market demand are formed. Only after processing, grading and pricing can the final products be formed for sale.

[0003] In the existing related technologies, visual data such as images of target tree seed-type cultural relics are captured or scanned by cameras and other equipment, and information such as the shape and outline of the target tree seed-type cultural relics is obtained by using operations such as image binarization segmentation. The specific dimensional features of the target tree seed-type cultural relics are calculated by manually designed feature extraction operators, so as to judge whether the target tree seed-type cultural relics can be paired based on feature similarity. However, the manually designed features only focus on the shallow features of one or some specific aspects of the image, and cannot fuse multiple features for matching. In addition, the manually designed features are limited to the number of tree seed-type cultural relics, and cannot be applied to the pairing and rapid search of large-scale multi-type target tree seed-type cultural relics, resulting in low efficiency and low accuracy in pairing between tree seed-type cultural relics. Summary of the invention

[0004] The present invention provides an automatic pairing system, method and electronic device for tree seed-type cultural relics, so as to solve the defects in the prior art that excessive reliance on manual design features leads to low pairing accuracy and slow efficiency between tree seed-type cultural relics, and realizes that an information processing device is used to determine the feature information of tree seed-type cultural relics in different dimensions to be paired, and a deep neural network pairing model is used to pair the tree seed-type cultural relics to be paired, so as to determine a set of paired tree seed-type cultural relics; then, the information processing device is used to output the placement position of each paired tree seed-type cultural relic in the paired tree seed-type cultural relic set in a collection and positioning storage cabinet, thereby improving the pairing accuracy and efficiency between tree seed-type cultural relics, and using the collection and positioning storage cabinet to quickly find tree seed-type cultural relics.

[0005] The present invention provides an automatic pairing system for tree seed-type collectibles, comprising a collection device, an information processing device, and a collection and positioning storage cabinet, wherein: The acquisition device is connected to the information processing device; The acquisition device is used to acquire characteristic images in various directions of at least one tree seed-type collectible to be paired; The information processing device is used to determine the characteristic information of each of the tree seed-type cultural relics to be paired in different dimensions based on each of the characteristic images; and generate the placement position of each of the tree seed-type cultural relics to be paired in the collection and positioning storage cabinet; The information processing device is also used to pair all the tree seed-type cultural relics to be paired based on the feature information of each of the tree seed-type cultural relics to be paired in different dimensions, the placement position corresponding to each of the tree seed-type cultural relics to be paired, and the pairing model, and determine the target placement position of each paired tree seed-type cultural relic in the paired tree seed-type cultural relic set in the collection and positioning storage cabinet and the evaluation index of at least one dimension corresponding to the paired tree seed-type cultural relic set; the pairing model is obtained by training the initial model of the deep neural network based on the feature information of the sample tree seed-type cultural relic; the target placement position is used to instruct the staff to take the paired tree seed-type cultural relic, and the evaluation index is used to characterize the grade of the paired tree seed-type cultural relic and the pricing of the paired tree seed-type cultural relic; The collection and positioning storage cabinet is used to place the tree seed-type collectibles to be paired according to the placement positions.

[0006] According to an automatic pairing system for tree seed-type collectibles provided by the present invention, the acquisition device is also used to acquire sample feature images of the sample tree seed-type collectibles in various directions; the information processing device is also used to determine sample feature information of the sample tree seed-type collectibles and sample pairing labels corresponding to the sample feature information based on the sample feature images; the sample feature information is input into the deep neural network initial model to obtain a model pairing result; based on the difference between the model pairing result and the sample pairing label, the deep neural network initial model is continuously optimized to obtain the pairing model.

[0007] According to an automatic pairing system for tree seed-type cultural relics provided by the present invention, the acquisition device is also used to acquire single feature images of a single tree seed-type cultural relic in various directions; the information processing device is also used to determine single feature information of different dimensions corresponding to the single tree seed-type cultural relic based on the single feature image; based on the single feature information, the feature information of each of the tree seed-type cultural relics to be paired, the placement position corresponding to each of the tree seed-type cultural relics to be paired and the pairing model, determine the placement position of a target tree seed-type cultural relic similar to the single tree seed-type cultural relic among the tree seed-type cultural relics to be paired in the acquisition and positioning storage cabinet and the evaluation index of at least one dimension corresponding to the target tree seed-type cultural relic.

[0008] According to an automatic pairing system for tree seed-like collectibles provided by the present invention, the collection equipment includes: a photographic light box, a main frame, a rotating table, a tree seed-like collectible stand, a bottom camera chassis, a top camera and a side camera, the main frame is placed at the bottom inside the photographic light box, the rotating table is placed at the bottom inside the main frame, the tree seed-like collectible stand is placed on the upper surface of the rotating table, the bottom camera chassis is placed between the rotating table and the tree seed-like collectible stand, the top camera is placed at the top inside the main frame, and the side camera is located on any side inside the main frame; the photographic light box is used for collection Gather characteristic images of the tree seed-like cultural relics to be recorded in various directions to provide a lighting environment; the tree seed-like cultural relics to be recorded include one or more of the tree seed-like cultural relics to be paired, the sample tree seed-like cultural relics, and the single tree seed-like cultural relics; the rotating table is used to drive the tree seed-like cultural relics to be recorded to rotate; the tree seed-like cultural relics support is used to place the tree seed-like cultural relics to be recorded; the bottom camera chassis is used to collect characteristic images of the tree seed-like cultural relics to be recorded in the bottom direction; the top camera is used to collect characteristic images of the tree seed-like cultural relics to be recorded in the top direction; the side camera is used to collect characteristic images of the tree seed-like cultural relics to be recorded in various directions on the horizontal plane.

[0009] According to an automatic pairing system for tree seed-type collectibles provided by the present invention, the collection and positioning storage cabinet includes: at least one drawer, a cabinet body number, and drawer numbers corresponding to each of the drawers; each of the drawers includes at least one square, and each of the squares includes a row number corresponding to the square and a column number corresponding to the square; the squares correspond one-to-one to the placement positions.

[0010] According to an automatic pairing system for tree seed-type cultural relics provided by the present invention, the automatic pairing system for tree seed-type cultural relics also includes: a printer; the printer is connected to the information processing device; the printer is used to print the evaluation index of at least one dimension corresponding to the paired tree seed-type cultural relics set and the evaluation index of at least one dimension corresponding to the target tree seed-type cultural relics.

[0011] The present invention provides an automatic pairing method for tree seed-type collectibles, which is applied to an information processing device in an automatic pairing system for tree seed-type collectibles. The method comprises: Acquire characteristic images in all directions of at least one tree seed-type collectible to be paired; Based on each of the characteristic images, characteristic information of each of the tree seed-type cultural relics to be paired in different dimensions is determined; and the placement position of each of the tree seed-type cultural relics to be paired in the collection and positioning storage cabinet is generated; Based on the characteristic information of each of the tree seed-type cultural relics to be paired in different dimensions, the corresponding placement positions and pairing models of each of the tree seed-type cultural relics to be paired, all the tree seed-type cultural relics to be paired are paired, and the target placement position of each paired tree seed-type cultural relic in the paired tree seed-type cultural relic set in the collection and positioning storage cabinet and the evaluation index of at least one dimension corresponding to the paired tree seed-type cultural relic set are determined; the pairing model is obtained by training the initial model of the deep neural network based on the characteristic information of the sample tree seed-type cultural relics; the target placement position is used to instruct the staff to take the paired tree seed-type cultural relic set, and the evaluation index is used to characterize the grade of the paired tree seed-type cultural relic and the pricing of the paired tree seed-type cultural relic.

[0012] According to an automatic pairing method for tree seed-type collectibles provided by the present invention, the training process of the pairing model includes: obtaining sample feature images of the sample tree seed-type collectibles in various directions; based on the sample feature images, determining sample feature information of the sample tree seed-type collectibles and sample pairing labels corresponding to the sample feature information; inputting the sample feature information into the deep neural network initial model to obtain a model pairing result; based on the difference between the model pairing result and the sample pairing label, continuously training the deep neural network initial model to obtain the pairing model.

[0013] An automatic pairing method for tree seed-type cultural relics provided by the present invention also includes: obtaining a single feature image of a single tree seed-type cultural relic in various directions; determining single feature information of different dimensions corresponding to the single tree seed-type cultural relic based on the single feature image; determining the placement position of a target tree seed-type cultural relic similar to the single tree seed-type cultural relic among the tree seed-type cultural relics to be paired in the collection and positioning storage cabinet and an evaluation index of at least one dimension corresponding to the target tree seed-type cultural relic based on the single feature information, the feature information of each of the tree seed-type cultural relics to be paired, the placement position corresponding to each of the tree seed-type cultural relics to be paired and the pairing model.

[0014] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the automatic pairing method for tree seed-type collectibles as described above is implemented.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for automatically pairing tree seed-type collectibles as described in any one of the above-mentioned methods is implemented.

[0016] The present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements the automatic pairing method for tree seed-type collectibles as described in any one of the above-mentioned methods.

[0017] The automatic pairing system, method and electronic device for tree seed-type cultural relics provided by the present invention collect feature images in various directions of at least one tree seed-type cultural relic to be paired through a collection device, and the information processing device determines the feature information of the tree seed-type cultural relic to be paired in different dimensions based on the feature images in various directions, thereby improving the diversity of features; the information processing device determines the tree seed-type cultural relic that can be paired among all the tree seed-type cultural relics to be paired based on the diversity features of the tree seed-type cultural relic to be paired, the placement positions of the tree seed-type cultural relic to be paired in the collection and positioning storage cabinet, and the pairing model of the deep neural network, and outputs the placement positions of the tree seed-type cultural relic that can be paired in the collection and positioning storage cabinet and the evaluation indicators of the tree seed-type cultural relic that can be paired. In this way, the operation speed is improved by utilizing the diversity features and the pairing model of the deep neural network, and the accuracy and efficiency of the pairing of the tree seed-type cultural relic are further improved. In addition, the collection and positioning storage cabinet can be used to quickly find the paired tree seed-type cultural relic, which is suitable for searching massive tree seed-type cultural relics. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 This is one of the structural schematic diagrams of the automatic pairing system for tree seed-type collectibles provided by the present invention.

[0020] Figure 2 It is a structural schematic diagram of the acquisition device provided by the present invention.

[0021] Figure 3 It is a structural schematic diagram of the collection and positioning storage cabinet provided by the present invention.

[0022] Figure 4 This is one of the flow charts of the automatic pairing method for tree seed-type collectibles provided by the present invention.

[0023] Figure 5 This is the second flow chart of the automatic pairing method for tree seed-type collectibles provided by the present invention.

[0024] Figure 6 This is the third flow chart of the automatic pairing method for tree seed-type collectibles provided by the present invention.

[0025] Figure 7 It is a schematic diagram of the flow chart of the pairing method of ornamental hemp walnuts according to the first embodiment provided by the present invention.

[0026] Figure 8It is a schematic diagram of the flow chart of the pairing method of ornamental hemp walnuts according to the second embodiment provided by the present invention.

[0027] Fig. 9 It is a schematic diagram of the flow chart of the pairing method of ornamental hemp walnuts according to the third embodiment provided by the present invention.

[0028] Fig.10 It is a schematic diagram of the user interface of the ornamental walnut pairing system provided by the present invention.

[0029] Fig.11 Schematic diagram of the user interface of the data collection option provided by the present invention.

[0030] Fig.12 is a schematic diagram of a user interface for starting pairing options provided by the present invention.

[0031] Fig.13 It is a schematic diagram of the pairing results in different directions provided by the present invention.

[0032] Fig.14 It is a schematic diagram of a user interface of a single matching option provided by the present invention.

[0033] Fig.15 It is a schematic diagram of the physical structure of the electronic device provided by the present invention.

[0034] Reference numerals: 100: Automatic pairing system for tree seed-type collectibles; 110: Collection equipment; 120: Information processing equipment; 130: Collection and positioning storage cabinet; 111: Photography light box; 112: Main frame; 113: Rotating table; 114: Tree seed-type collectibles stand; 115: Bottom camera chassis; 116: Top camera; 117: Side camera; 131: Drawer; 132: Block. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution 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, not 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.

[0036] Tree seed-type wenwan can be the raw material for making products or the product itself. For example, tree seed-type wenwan walnuts are products themselves. Tree seed-type wenwan can also be used as raw materials to make bracelets, Buddhist beads and other products. In the early stage of the development of wenwan, after the tree seed-type wenwan is harvested, professional matchers are needed to manually screen and match, and finally form paired finished products or classified bracelet materials and hand-held materials that meet market demand. After processing, grading and pricing, the final product can be sold. It belongs to the category of agricultural product deep processing that relies heavily on manual labor. However, there are many varieties of tree seed-type wenwan products with strange shapes, ever-changing textures, and sizes. The final market value of each category of tree seed-type wenwan products is closely related to the size, similarity, texture matching degree, and prominence of variety characteristics of the finished product. Similar to the complexity of antique wenwan appraisal, tree seed-type wenwan requires a high degree of specialization. Manual screening and matching alone requires high comprehensive qualities such as industry literacy, operating experience, and aesthetic factors of the screening and matching personnel. The efficiency of manual screening and matching is generally low, so that the quality of the finished products is uneven. The heavy reliance on manual labor makes it difficult for this industry to achieve industrialization and large-scale production.

[0037] There are a small number of methods on the market that use computer vision methods to obtain the features of tree seed-type cultural relics, replacing the manual screening and matching methods. They use cameras and other equipment to capture or scan images and other visual data of target tree seed-type cultural relics, use image binarization and segmentation and other operations to obtain information such as the shape and outline of the target tree seed-type cultural relics, and use manually designed feature extraction operators to calculate the specific dimensional features of the target tree seed-type cultural relics. They then judge whether the target tree seed-type cultural relics can be matched based on feature similarity. However, manually designed features only focus on shallow features of one or some specific aspects of the image, and cannot fuse multiple features for matching. In addition, manually designed features are limited to the number of tree seed-type cultural relics, and cannot be applied to large-scale matching and rapid search of multiple types of target tree seed-type cultural relics, resulting in low efficiency and accuracy in matching between tree seed-type cultural relics.

[0038] In addition, existing computer vision methods do not have reasonable storage management equipment and fast access methods that are suitable for large-scale and multi-category target tree seed-type cultural relics.

[0039] Combine the following Figure 1-Figure 3 The automatic pairing system for tree seed-type collectibles of the present invention is described. The automatic pairing system for tree seed-type collectibles is suitable for automatic screening and pairing of any tree seed-type collectibles and products, including but not limited to hemp walnuts, iron walnuts, elm walnuts, lantern walnuts, etc.

[0040] Figure 1 This is one of the structural schematic diagrams of the automatic pairing system for tree seed-type collectibles provided by the present invention. Figure 1As shown, the automatic pairing system 100 for tree seed-type collectibles includes: a collection device 110, an information processing device 120, and a collection and positioning storage cabinet 130, wherein: The acquisition device 110 is connected to the information processing device 120; The acquisition device 110 is used to acquire characteristic images in various directions of at least one tree seed-type collectible to be paired; The information processing device 120 is used to determine the feature information of each of the tree seed-type cultural relics to be paired in different dimensions based on each of the feature images; and generate the placement position of each of the tree seed-type cultural relics to be paired in the collection and positioning storage cabinet 130; The information processing device 120 is further used to pair all the tree seed-type cultural relics to be paired based on the feature information of each of the tree seed-type cultural relics to be paired in different dimensions, the placement position corresponding to each of the tree seed-type cultural relics to be paired, and the pairing model, and determine the target placement position of each paired tree seed-type cultural relic in the paired tree seed-type cultural relic set in the collection and positioning storage cabinet 130 and the evaluation index of at least one dimension corresponding to the paired tree seed-type cultural relic set; the pairing model is obtained by training the initial model of the deep neural network based on the feature information of the sample tree seed-type cultural relic; the target placement position is used to instruct the staff to take the paired tree seed-type cultural relic, and the evaluation index is used to characterize the grade of the paired tree seed-type cultural relic and the pricing of the paired tree seed-type cultural relic; The collection and positioning storage cabinet 130 is used to place the tree seed-type collectibles to be paired according to the placement positions.

[0041] Here, the tree seed-type collectibles to be paired include but are not limited to raw seeds such as hemp walnuts, iron walnuts, catalpa walnuts, lantern walnuts, etc. Among them, the tree seed-type collectibles can be used directly as products, or they can be processed into other products such as bracelets and Buddhist beads.

[0042] It should be noted that the time required for the collection device 110 to collect a single tree seed-type collectible to be paired may be in the order of seconds.

[0043] Here, the characteristic images in various directions may refer to the six-sided color characteristic images of the front, back, left, right, top and bottom of the tree seed-type collectibles to be paired.

[0044] The information processing device 120 may include a computer, a server, a cloud, etc. The information processing device 120 can store data and perform calculations. When the information processing device 120 is a computer, it is a local collection, storage, and calculation. The local computer serves as a data storage and calculation system. The data of the tree seed-type cultural relics to be paired is transmitted from the collection device 110 to the local computer. The local computer calls a high-performance graphics processing unit (GPU) processor to directly calculate the deep-level features of the tree seed-type cultural relics to be paired, and performs screening and matching tasks; when the information processing device 120 is a cloud server, local collection and cloud server storage and calculation are performed. The local computer only serves as a collection status display terminal. After the data of the tree seed-type cultural relics to be paired is transmitted from the collection device 110 to the local computer, it is uploaded to the cloud server by the local computer. The cloud server calls a processor such as a GPU to calculate the deep-level features of the target tree seed-type cultural relics, and performs screening and matching tasks.

[0045] The collection and positioning storage cabinet 130 can be a cabinet of any shape, material or size for positioning.

[0046] Here, the characteristic information of different dimensions includes but is not limited to the size data of tree seed-type collectibles (edge ​​width data, belly width data, top-to-bottom height data), color temperature and color number data, variety characteristic data, texture coarseness and distribution direction data, pile type data and other characteristic data. Among them, the edge width data, belly width data and top-to-bottom height data can be obtained through pixel calculation of the collected image, the color temperature and color number data are obtained by comparing the six-sided color image parameters with the imaging data of the variety in the same system, the variety characteristic data value is obtained by comparing the imaging data of the variety in the same model system, and the texture coarseness and distribution direction data and pile type data are obtained by comparing the variety characteristic parameters captured by the deep learning neural network model of the variety in the same system.

[0047] Here, the method for determining the feature information can be obtained through a calculation model; it can also be obtained through a pre-set mapping relationship table between feature images and feature information; it can also be obtained based on an image extraction model, which is not limited in the present invention.

[0048] Here, the placement positions of each paired tree seed-type collectible can be placed in order in the collection and positioning storage cabinet 130, for example, the first collected tree seed-type collectible is placed in the first position in the collection and positioning storage cabinet 130, and the second collected tree seed-type collectible is placed in the second position in the collection and positioning storage cabinet 130; it can also be determined according to a pre-set mapping relationship table between tree seed-type collectibles and placement positions, and the present invention is not limited to this.

[0049] Here, pairing means finding two or more similar tree seed-type collectibles among all tree seed-type collectibles. When pairing, it is necessary to integrate multiple feature information to calculate the similarity. The higher the similarity, the higher the pairing rate; the lower the similarity, the lower the pairing rate. The method for calculating the similarity can use Euclidean distance, cosine similarity, text matching model, etc., and the present invention does not limit this.

[0050] Here, the evaluation index may include one or more of variety information, size information, similarity information, variety display information, and automatic grading information.

[0051] The paired tree seed-type collection of cultural relics may include two similar tree seed-type cultural relics or more than two similar tree seed-type cultural relics. All the tree seed-type cultural relics to be paired may be paired into one or more paired tree seed-type cultural relics sets. It should be noted that all the tree seed-type cultural relics to be paired in the database will be calculated through the trained pairing model for deep-level features, and compared. The matrix will be divided according to the placement position and similarity, and finally two or more most similar placement positions will be generated according to the pairing weight threshold of each variety. The two results generated here can be considered to complete the traditional tree seed-type cultural relics pairing requirements. A pair of walnuts is two walnuts (tree seed-type cultural relics). Multiple results can be selected according to specific needs. For example, 14 of the most similar tree seed-type cultural relics can be selected as a set of raw materials for hand strings, and 108 of the most similar tree seed-type cultural relics can be selected as a set of raw materials for Buddhist beads. The paired product parameters here will reflect multi-dimensional information in the system, such as similarity information, size information, variety signature feature information, etc. The variety signature information mentioned here will serve as an important parameter basis for the final automatic grading. Taking the tree seed type of walnut as an example, it is not the case that two walnuts with high similarity can be paired to form a good paired product and form a high market value. In addition to similarity, the value of walnuts also includes the pile type of the pair, the degree of prominence of its variety characteristics, its texture direction and texture fullness, its variety scarcity, its skin oiliness and initial color, its tree-dropping time and maturity, etc. For example: the variety of Sizuolou Lion Head, which is native to Pinggu District, Beijing, has the variety imaging characteristics of wide-rimmed large ears, straight waterfall bumps and flowing water patterns, mountain-shaped tips, and chrysanthemum bottoms. Therefore, the grading definition factors for this walnut variety should include the larger size, positive pile, high similarity, etc. in the variety, and the dominant characteristics of this variety should be more obvious. If all these factors are present, this pair of walnuts belongs to the category of fine products and will have a higher market value. Therefore, when grading the paired results of each variety, in addition to ensuring the accuracy of the pairing, the system also needs to automatically adapt to the dominant characteristics of the variety.

[0052] Specifically, after the acquisition device 110 acquires the characteristic image of the tree seed-type cultural relics to be paired, the characteristic image is transmitted to the information processing device 120. After the information processing device 120 determines the characteristic information based on the characteristic image, the characteristic information of each tree seed-type cultural relics to be paired is saved in a local or cloud database. While calculating the characteristic information of each tree seed-type cultural relics to be paired, the information processing device 120 displays the placement position of the tree seed-type cultural relics to be paired in the acquisition and positioning storage cabinet 130 on the display screen. The staff places the tree seed-type cultural relics to be paired in the corresponding position in the acquisition and positioning storage cabinet 130 according to the placement position. Repeat the acquisition, calculation, and generation of placement positions for each tree seed-type cultural relics to be paired until the staff places all the tree seed-type cultural relics to be paired in the corresponding positions in the acquisition and positioning storage cabinet 130. At this point, the input work is completed. Then, the information processing device 120 loads the pairing model to extract features and screen similarities of the paired tree seed-type cultural relics, generates the target placement position and multiple evaluation indicators of each paired tree seed-type cultural relic in the collection and positioning storage cabinet 130 in the paired tree seed-type cultural relic set, and displays the target placement position and the similarity information of the paired tree seed-type cultural relic, the variety feature visualization information, and the automatic grading information and other multi-dimensional evaluation indicators on the display screen; the staff takes the corresponding number from the collection and positioning storage cabinet 130 according to the target placement position to form a tree seed-type cultural relic product, and the staff classifies and prices the product according to the evaluation indicators. Here, the display can be a display on the information processing device 120 or a separate display.

[0053] It should be noted that when the pairing similarity calculation and screening is performed by loading the pairing model, the high-performance graphics processing unit (GPU) is used for operation, and the operation speed is fast. The deep neural network model is used to automatically set the feature priority according to the characteristics of the variety, and automatically complete the position display and evaluation indicators (such as similarity and grading) display according to the characteristics of the variety. It should be understood that the grading in the evaluation index is the system's preliminary grading of the paired tree seed-type cultural relics. Later, the staff can comprehensively determine the final grading of the tree seed-type cultural relics to be paired based on multiple evaluation indicators.

[0054] In an embodiment of the present invention, the acquisition device 110 acquires feature images in various directions of at least one tree seed-type cultural relic to be paired, and the information processing device 120 determines feature information of the tree seed-type cultural relic to be paired in different dimensions based on the feature images in various directions, thereby improving the diversity of features; the information processing device 120 determines the tree seed-type cultural relic that can be paired among all the tree seed-type cultural relic to be paired based on the diversity features of the tree seed-type cultural relic to be paired, the placement positions of the tree seed-type cultural relic to be paired in the acquisition and positioning storage cabinet 130, and the pairing model of the deep neural network, and outputs the placement positions of the tree seed-type cultural relic that can be paired in the acquisition and positioning storage cabinet 130 and the evaluation indicators of the tree seed-type cultural relic that can be paired. In this way, the operation speed is improved by utilizing the diversity features and the pairing model of the deep neural network, thereby further improving the accuracy and efficiency of pairing the tree seed-type cultural relic. In addition, the acquisition and positioning storage cabinet 130 can be used to quickly find paired tree seed-type cultural relic, which is suitable for searching massive tree seed-type cultural relic.

[0055] Furthermore, the acquisition device 110 is also used to acquire sample feature images of the sample tree seed-type collectibles in various directions; The information processing device 120 also determines the sample feature information of the sample tree seed-type cultural relics and the sample pairing labels corresponding to the sample feature information based on the sample feature image; inputs the sample feature information into the deep neural network initial model to obtain a model pairing result; and continuously optimizes the deep neural network initial model based on the difference between the model pairing result and the sample pairing label to obtain the pairing model.

[0056] Here, the sample tree seed-type collectibles may include different varieties of tree seed-type collectibles, and the sample feature information may include one or more of size data (including side width data, belly width data, top-to-bottom height data, etc.), color temperature and color number data, variety feature data, texture thickness and distribution direction data, pile type data, and other feature data, and the sample pairing label may include: evaluation indicators of paired sample tree seed-type collectibles, such as similarity information, grading information, etc. A sample training data set is formed by obtaining the sample feature information and sample pairing labels of the sample tree seed-type collectibles.

[0057] Here, the initial deep neural network model can be built by the staff or downloaded from the cloud or server.

[0058] It should be noted that the purpose of training the initial model of the deep neural network is to minimize the difference between the model pairing results and the sample pairing labels, and to obtain the pairing model by continuously iteratively optimizing the difference between the model pairing results and the sample pairing labels.

[0059] In an embodiment of the present invention, the training of the pairing model only needs to be done once, and it can be used repeatedly in the actual pairing method process. In order to improve the performance of the model, the model can also be trained simultaneously with the data of multiple varieties of tree seed-type cultural materials, or the model can be trained incrementally in batches. The model will grow iteratively as the number of trained varieties gradually increases. For example, when the pairing model only trains the to-be-screened paired variety a, the software system that loads the pairing model of the variety can only automatically screen and pair the a variety. After continuing to train the variety b on the basis of the training of the pairing model of variety a, the software system that loads the composite variety pairing model can complete the automatic screening and pairing of both variety a and variety b. Therefore, having iterative growth attributes is a significant feature of the pairing model. The pairing model only needs to be trained once and can be copied, and all systems that load the pairing model can realize the automatic screening and pairing of the trained varieties, which has great exploration significance for the cloud computing system based on the pairing model. When the training work of the pairing model and the software system that loads the pairing model perform calculations, the calculation work can be completed based on the GPU, and its calculation speed is significantly improved compared with the calculation speed of the traditional central processing unit (CPU).

[0060] In an embodiment of the present invention, by continuously training the initial model of the deep neural network, the model is enabled to autonomously learn the characteristics of tree seed-type cultural relics, perform similarity matching, and achieve end-to-end, thereby improving the robustness, generalization ability, accuracy and computing speed of the pairing model.

[0061] Furthermore, the acquisition device 110 is also used to acquire a single characteristic image of a single tree seed-type collectible in various directions; The information processing device 120 is also used to determine the single feature information of different dimensions corresponding to the single tree seed-type cultural relics based on the single feature image; based on the single feature information, the feature information of each of the tree seed-type cultural relics to be paired, the placement position corresponding to each of the tree seed-type cultural relics to be paired and the pairing model, determine the placement position of the target tree seed-type cultural relics similar to the single tree seed-type cultural relics among the tree seed-type cultural relics to be paired in the collection and positioning storage cabinet 130 and the evaluation index of at least one dimension corresponding to the target tree seed-type cultural relics.

[0062] Here, the evaluation index may include one or more of variety information, size information, similarity information, variety display information, and automatic grading information.

[0063] It should be noted that, in addition to pairing multiple tree seed-type cultural relics with each other, the automatic pairing system of the tree seed-type cultural relics of the present invention can also find one or more tree seed-type cultural relics that are similar to a single tree seed-type cultural relic among multiple tree seed-type cultural relics. For example, a customer has a single tree seed-type cultural relic and needs to match the tree seed-type cultural relic with the most similar tree seed-type cultural relic to form a pair.

[0064] Specifically, the acquisition device 110 acquires feature images of a single tree seed-type cultural relic in different directions, determines feature information in different dimensions based on the feature images, and the information processing device 120 loads the pairing model to find a target tree seed-type cultural relic that is similar to the single tree seed-type cultural relic from multiple tree seed-type cultural relic to be paired, and outputs the placement position of the target tree seed-type cultural relic in the acquisition and positioning storage cabinet 130 and the evaluation index corresponding to the target tree seed-type cultural relic.

[0065] It should be noted that when a single tree seed-type collectible is paired, it is not necessary to generate a placement position for the collection and positioning storage cabinet 130 for the single tree seed-type collectible.

[0066] It should be noted that if there is already a database of tree seed-type cultural relics to be matched, when matching a single tree seed-type cultural relic, the match can be made directly from the database of tree seed-type cultural relics to be matched; if there is no database of tree seed-type cultural relics to be matched, it is necessary to first enter the characteristic information of the tree seed-type cultural relics to be matched to form a database of tree seed-type cultural relics to be matched, and then perform the matching work.

[0067] In the embodiment of the present invention, the automatic pairing system for tree seed-type collectibles can solve the matching task of a single tree seed-type collectible with an existing database of tree seed-type collectibles, thereby improving the wide applicability of the automatic pairing system for tree seed-type collectibles.

[0068] Further, the acquisition device 110 includes: a photography light box, a main frame, a rotating table, a tree seed type plaything stand, a bottom camera chassis, a top camera and a side camera; The main frame is placed at the bottom of the photography light box, the rotating table is placed at the bottom of the main frame, the tree seed-like ornaments stand is placed on the upper surface of the rotating table, the bottom camera chassis is placed between the rotating table and the tree seed-like ornaments stand, the top camera is placed at the top of the main frame, and the side camera is located on any side of the main frame; The photography light box is used to provide a lighting environment for collecting characteristic images in various directions of the tree seed-type cultural relics to be recorded; the tree seed-type cultural relics to be recorded include one or more of the tree seed-type cultural relics to be paired, the sample tree seed-type cultural relics, and the single tree seed-type cultural relics; The rotating table is used to drive the tree seed-like collectibles to be recorded to rotate; The tree seed type collectibles support is used to place the tree seed type collectibles to be recorded; The bottom camera chassis is used to collect the characteristic image of the tree seed-type collectibles to be recorded in the bottom direction; The top camera is used to collect the characteristic image of the tree seed-type collectibles to be recorded in the top direction; The side camera is used to collect characteristic images of the tree seed-type collectibles to be recorded in various directions on the horizontal plane.

[0069] It should be noted that the photography light box can also perform supplementary lighting and color restoration for the entire acquisition device. The main frame realizes the high integration of the acquisition device.

[0070] Here, the bottom camera chassis includes a bottom camera, which collects characteristic images of the tree seed-type cultural relics to be recorded in the bottom direction, and also includes the integration of components such as a power supply device, a main circuit board, and a sub-program storage medium.

[0071] For example, Figure 2 is a schematic diagram of the structure of the acquisition device 110 provided by the present invention, such as Figure 2 As shown, the acquisition device 110 includes a photographic light box 111, a main frame 112, a rotating platform 113, a tree seed type ornament stand 114, a bottom camera chassis 115, a top camera 116 and a side camera 117. In the process of acquiring characteristic images in various directions, the rotating platform 113 will rotate to enable the side camera 117 to acquire characteristic images of the tree seed type ornament in four directions: front, back, left and right.

[0072] In the embodiment of the present invention, the collection device 110 is set to be open, with a fast collection speed (seconds) and a wide dimension of collected data, which is conducive to the later generation of position coordinate information and the construction of a detailed database. It is particularly suitable for the screening and matching of massive tree seed-type cultural relics, and has high collection operation efficiency.

[0073] Furthermore, the collection and positioning storage cabinet 130 includes: at least one drawer, a cabinet number, and drawer numbers corresponding to each of the drawers; each of the drawers includes at least one square, and each of the squares includes a row number corresponding to the square and a column number corresponding to the square; the squares correspond one-to-one to the placement positions.

[0074] It should be understood that one square corresponds to a placement position. The placement position of tree seed-type collectibles can be based on cabinet number-drawer number-grid row number in the drawer-grid column number in the drawer. For example, A-1-A-1 means that the tree seed-type collectibles are placed in the first drawer of cabinet A in the Ath row and the 1st column.

[0075] For example, Figure 31 is a schematic diagram of the structure of the collection and positioning storage cabinet 130 provided by the present invention. Figure 3 As shown, the collection and positioning storage cabinet 130 includes a drawer 131 and a grid 132, and the A on the collection and positioning storage cabinet 130 represents the number of the cabinet, and each drawer has a corresponding drawer number, such as 1-20, and each drawer is divided into a number of squares by the grid, and each square has a row number (such as A, B, C, D, E, F, G) and a list number (such as 1, 2, 3, 4, 5, 6, 7). The generated placement position is expressed as cabinet code-drawer code-row code-column code. For example, the longitude and latitude positioning information automatically generated by the system when the target tree seed-type cultural relics are entered is A-1-A-1, and the code indicates that the tree seed-type cultural relics should be placed in cabinet A, drawer 1, first row, first column.

[0076] In an embodiment of the present invention, the grid design of the collection and positioning storage cabinet 130 allows the staff to quickly pick up the matched target tree seed-type cultural relics according to the placement position displayed in the result after the matching result is displayed, making the entire matching process simple and easy to operate, eliminating the trouble of searching for the target tree seed-type cultural relics, realizing a fully closed-loop workflow for the matching process, and avoiding the problem of searching for massive tree seed-type cultural relics.

[0077] Furthermore, it also includes: a printer 140; the printer 140 is connected to the information processing device 120; The printer is used to print the evaluation index of at least one dimension corresponding to the paired tree seed type cultural relics set and the evaluation index of at least one dimension corresponding to the target tree seed type cultural relics.

[0078] Here, the printer may be any suitable printer, such as a flexibly configurable small printer, a laser printer, a color printer, etc.

[0079] It should be noted that the staff can customize the label information printed by the printer, and the label information can be displayed according to user needs, such as year information, variety information, size information, similarity information, grading information, dealer information, etc.

[0080] In the embodiment of the present invention, the evaluation index is printed by a printer to assist the staff in the grading and pricing of finished products, so that the products are more formalized and the processes are more standardized.

[0081] Figure 4 It is one of the flow charts of the automatic pairing method of tree seed-type cultural relics provided by the present invention, and is applied to the information processing equipment in the automatic pairing system of tree seed-type cultural relics, such as Figure 4 As shown, the method includes: Step 401: Obtain feature images in various directions of at least one tree seed-type collectible to be paired.

[0082] Step 402: based on each of the characteristic images, determine the characteristic information of each of the tree seed-type cultural relics to be paired in different dimensions; and generate the placement position of each of the tree seed-type cultural relics to be paired in the collection and positioning storage cabinet.

[0083] Step 403: Pair all the tree seed-type cultural relics to be paired based on the feature information of each of the tree seed-type cultural relics to be paired in different dimensions, the placement position and pairing model corresponding to each of the tree seed-type cultural relics to be paired, and determine the target placement position of each paired tree seed-type cultural relic in the paired tree seed-type cultural relic set in the collection and positioning storage cabinet and the evaluation index of at least one dimension corresponding to the paired tree seed-type cultural relic set.

[0084] Here, the pairing model is obtained by training the initial model of the deep neural network based on the feature information of the sample tree seed-type cultural relics; the target placement position is used to instruct the staff to take the paired tree seed-type cultural relics, and the evaluation index is used to characterize the grade of the paired tree seed-type cultural relics and the pricing of the paired tree seed-type cultural relics.

[0085] Figure 5 This is a second flow chart of the automatic pairing method for tree seed-type collectibles provided by the present invention. Figure 5 As shown, the training process of the pairing model includes: Step 501, obtaining sample feature images of the sample tree seed-type collectibles in various directions; Step 502: Based on the sample feature image, determine the sample feature information of the sample tree seed-type collectibles and the sample pairing label corresponding to the sample feature information; Step 503: input the sample feature information into the initial model of the deep neural network to obtain a model pairing result; Step 504: Based on the difference between the model pairing result and the sample pairing label, continuously train the deep neural network initial model to obtain the pairing model.

[0086] Figure 6 This is the third flow chart of the automatic pairing method for tree seed-type collectibles provided by the present invention, such as Figure 6 As shown, the method also includes: Step 601, obtaining a single characteristic image of a single tree seed-type collectible in various directions; Step 602: Based on the single feature image, determine the single feature information of different dimensions corresponding to the single tree seed-type collectible; Step 603, based on the single feature information, the feature information of each of the tree seed-type cultural relics to be paired, the placement position corresponding to each of the tree seed-type cultural relics to be paired and the pairing model, determine the placement position of the target tree seed-type cultural relics among the tree seed-type cultural relics to be paired that are similar to the single tree seed-type cultural relics in the collection and positioning storage cabinet 130 and the evaluation index of at least one dimension corresponding to the target tree seed-type cultural relics.

[0087] The following are three embodiments of the automatic pairing method for tree seed-type collectibles provided by the present invention.

[0088] It should be noted that the tree seeds to be paired in the three embodiments are hemp walnuts.

[0089] Embodiment 1 is applicable to the case where the species to be paired is a new species and needs to be trained with a deep neural network model before pairing with tree seeds, thus realizing a closed loop from collection, storage, calculation, and retrieval.

[0090] Figure 7 : is a flow chart of a pairing method for playing hemp walnuts according to the first embodiment of the present invention, such as Figure 7 As shown, the pairing method of the Wenwan hemp walnuts comprises the following steps: S701: Use the collection equipment to collect the multi-dimensional feature information of the sample Wenwan hemp walnut.

[0091] S702: Construct and train a pairing model for feature extraction of Wenwan hemp walnuts based on deep learning.

[0092] S703: Using a collection device to collect multi-dimensional feature information of the hemp walnuts to be paired and save it to a database.

[0093] S704: After the data of each hemp walnut to be paired is collected, a placement position is generated and the data is stored in the collection and positioning storage cabinet in sequence.

[0094] S705: Use the trained pairing model to calculate features and similarities to complete pairing.

[0095] S706: The staff takes the paired hemp walnuts to be played with according to their placement shown in the system and prices them based on multi-dimensional information.

[0096] S707: The staff uses the system to link the label printing equipment, customizes the variety, size, similarity and other information and prints it to complete the productization.

[0097] Embodiment 2 is applicable to the system that already has a mature deep neural network model covering multiple varieties. It can skip the deep learning training procedure of the neural network model and directly perform the collection and pairing work, which simplifies the training process and printing options, and highlights the closed-loop workflow and high efficiency.

[0098] Figure 8 : is a flow chart of a pairing method for playing hemp walnuts according to the second embodiment of the present invention, such as Figure 8 As shown, the pairing method of the Wenwan hemp walnuts comprises the following steps: S801: Use a collection device to collect multi-dimensional feature information of the hemp walnuts to be paired and save it to a database.

[0099] S802: After the data of each hemp walnut to be paired is collected, a placement position is generated and the data is stored in the collection and positioning storage cabinet in sequence.

[0100] S803: Use the trained pairing model to calculate features and similarities to complete pairing.

[0101] S804: The staff takes the paired hemp walnuts for playing according to the placement of the paired hemp walnuts displayed on the system and prices them according to the multi-dimensional information.

[0102] Embodiment 3 is applicable to the system that already has a formed neural network model covering multiple varieties. The staff currently holds a single Wenwan walnut a and needs to find a hemp walnut b in the database of the same variety that can form the best pair with the hemp walnut a, so as to solve the matching task of the specified single Wenwan hemp walnut with the existing Wenwan hemp walnut database.

[0103] Fig. 9 : is a flow chart of the pairing method of the Wenwan hemp walnuts according to the third embodiment of the present invention, such as Fig. 9 As shown, the pairing method of the walnuts includes the following steps: S901: Use a collection device to collect multi-dimensional feature information of the hemp walnuts to be paired and save it to a database.

[0104] S902: After the data of each hemp walnut to be paired is collected, a placement position is generated and the data is stored in the collection and positioning storage cabinet in sequence.

[0105] S903: Using a collection device to collect multi-dimensional feature information of a single hemp walnut to be paired.

[0106] S904: Use the trained pairing model to calculate features and similarities to find the best match.

[0107] S905: The staff takes the target Wenwan hemp walnuts according to the placement of the target Wenwan hemp walnuts displayed by the system and prices them according to the multi-dimensional information.

[0108] Among them, steps S901~S902 can be performed only once, and subsequently, for matching other multiple single ornamental hemp walnuts, only steps S903~S905 need to be executed.

[0109] It should be noted that the system provided by the present invention will eventually be implemented through a user interface. Fig.10 : is a user interface diagram of the walnut pairing system provided by the present invention, such as Fig.10 As shown, the user interface includes multiple setting options, such as data collection, start pairing, single matching, general settings, etc. Different setting options realize different functions. Among them, the data collection option can realize the function of collecting data, the start pairing option can realize the function of matching the walnuts to be paired, and the single pairing option can realize the function of pairing a single walnut.

[0110] Fig.11 is a user interface schematic diagram of the data collection option provided by the present invention, such as Fig.11 As shown, the user interface of data collection includes variety settings, collection box settings, information query, images of Wenwan walnuts from different perspectives, the three-dimensional size of Wenwan walnuts, the position information (placing position) of Wenwan walnuts in the collection and positioning storage cabinet 130, and a progress bar for information collection.

[0111] Fig.12 is a user interface schematic diagram of the start pairing option provided by the present invention, such as Fig.12 As shown, the user interface for starting pairing includes options such as variety selection, size selection, data query, and after pairing, the variety, size, location information (placement), similarity, level, and printing of the pairing results. Subsequent staff can comprehensively evaluate the level and price of the paired tree seed-type collectibles based on a variety of information.

[0112] Fig.13 is a schematic diagram of the pairing results in different directions provided by the present invention, such as Fig.13 As shown, the six columns of images respectively represent the similarity results in six directions: top, bottom, front, back, left, and right, and also include comprehensive similarity, size, color, position and other information.

[0113] Fig.14 is a schematic diagram of a user interface for a single matching option provided by the present invention, such as Fig.14 As shown, in the single matching user interface, the user collects images of a single Wenwan walnut in six directions and displays the three-dimensional size of the single Wenwan walnut through operations, clicks to start matching, and after the matching is completed, the three-dimensional size and position information of the matched Wenwan walnut are displayed.

[0114] Fig.15 is a schematic diagram of the physical structure of the electronic device provided by the present invention, such as Fig.15As shown, the electronic device may include: a processor (processor) 1510 , a communication interface (Communications Interface) 1520 , a memory (memory) 1530 and a communication bus 1540 , wherein the processor 1510 , the communication interface 1520 , and the memory 1530 communicate with each other through the communication bus 1540 . The processor 1510 can call the logic instructions in the memory 1530 to execute the automatic pairing method of tree seed-type cultural relics, which includes: obtaining feature images in various directions of at least one tree seed-type cultural relic to be paired; based on each of the feature images, determining the feature information of each of the tree seed-type cultural relics to be paired in different dimensions; and generating the placement position of each of the tree seed-type cultural relics to be paired in the collection and positioning storage cabinet 130; based on the feature information of each of the tree seed-type cultural relics to be paired in different dimensions, the placement position corresponding to each of the tree seed-type cultural relics to be paired, and the pairing model, pairing all the tree seed-type cultural relics to be paired, and determining the target placement position of each paired tree seed-type cultural relic in the paired tree seed-type cultural relic set in the collection and positioning storage cabinet 130 and the evaluation index of at least one dimension corresponding to the paired tree seed-type cultural relic set; the pairing model is obtained by training the initial model of the deep neural network based on the feature information of the sample tree seed-type cultural relic; the target placement position is used to instruct the staff to take the target tree seed-type cultural relic, and the evaluation index is used to characterize the pricing of the target tree seed-type cultural relic.

[0115] In addition, the logic instructions in the above-mentioned memory 1530 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0116] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the automatic pairing method of tree seed-type cultural relics provided by the above-mentioned methods, which method includes: obtaining a feature image in various directions of at least one tree seed-type cultural relic to be paired; based on each of the feature images, determining the feature information of each of the tree seed-type cultural relics to be paired in different dimensions; and generating a placement position of each of the tree seed-type cultural relics to be paired in the collection and positioning storage cabinet 130; based on the placement position of each of the tree seed-type cultural relics to be paired in different dimensions, The feature information of the same dimension, the placement position corresponding to each of the tree seed-type cultural relics to be paired and the pairing model are used to pair all the tree seed-type cultural relics to be paired, and the target placement position of each paired tree seed-type cultural relic in the paired tree seed-type cultural relic set in the collection and positioning storage cabinet 130 and the evaluation index of at least one dimension corresponding to the paired tree seed-type cultural relic set are determined; the pairing model is obtained by training the initial model of the deep neural network based on the feature information of the sample tree seed-type cultural relics; the target placement position is used to instruct the staff to take the target tree seed-type cultural relic, and the evaluation index is used to characterize the pricing of the target tree seed-type cultural relic.

[0117] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the automatic pairing method of tree seed-type cultural relics provided by the above-mentioned methods, the method comprising: obtaining a characteristic image in various directions of at least one tree seed-type cultural relic to be paired; based on each of the characteristic images, determining the characteristic information of each of the tree seed-type cultural relic to be paired in different dimensions; and generating a placement position of each of the tree seed-type cultural relic to be paired in the collection and positioning storage cabinet 130; based on the characteristic information of each of the tree seed-type cultural relic to be paired in different dimensions, The placement positions and pairing models corresponding to tree seed-type cultural relics are used to pair all the tree seed-type cultural relics to be paired, and the target placement positions of each paired tree seed-type cultural relic in the paired tree seed-type cultural relic set in the collection and positioning storage cabinet 130 and the evaluation index of at least one dimension corresponding to the paired tree seed-type cultural relic set are determined; the pairing model is obtained by training the initial model of the deep neural network based on the feature information of the sample tree seed-type cultural relics; the target placement position is used to instruct the staff to take the target tree seed-type cultural relic, and the evaluation index is used to characterize the pricing of the target tree seed-type cultural relic.

[0118] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0119] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0120] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An automatic pairing system for tree seed-type collectibles, characterized in that: include: Collection equipment, information processing equipment and collection and positioning storage cabinets, including: The acquisition device is connected to the information processing device; The acquisition device is used to acquire characteristic images in various directions of at least one tree seed-type collectible to be paired; The information processing device is used to determine the characteristic information of each of the tree seed-type cultural relics to be paired in different dimensions based on each of the characteristic images; and generate the placement position of each of the tree seed-type cultural relics to be paired in the collection and positioning storage cabinet; The information processing device is also used to pair all the tree seed-type cultural relics to be paired based on the feature information of each of the tree seed-type cultural relics to be paired in different dimensions, the placement position corresponding to each of the tree seed-type cultural relics to be paired, and the pairing model, and determine the target placement position of each paired tree seed-type cultural relic in the paired tree seed-type cultural relic set in the collection and positioning storage cabinet and the evaluation index of at least one dimension corresponding to the paired tree seed-type cultural relic set; the pairing model is obtained by training the initial model of the deep neural network based on the feature information of the sample tree seed-type cultural relic; the target placement position is used to instruct the staff to take the paired tree seed-type cultural relic, and the evaluation index is used to characterize the grade of the paired tree seed-type cultural relic and the pricing of the paired tree seed-type cultural relic; The collection and positioning storage cabinet is used to place the tree seed-type collectibles to be paired according to the placement positions.

2. The automatic pairing system for tree seed-type collectibles according to claim 1 is characterized in that: The acquisition device is also used to acquire sample feature images of the sample tree seed-type collectibles in various directions; The information processing device is further used to determine the sample characteristic information of the sample tree seed-type cultural relics and the sample pairing label corresponding to the sample characteristic information based on the sample characteristic image; The sample feature information is input into the deep neural network initial model to obtain a model pairing result; based on the difference between the model pairing result and the sample pairing label, the deep neural network initial model is continuously optimized to obtain the pairing model.

3. The automatic pairing system for tree seed-type collectibles according to claim 1 is characterized in that: The acquisition device is also used to acquire a single characteristic image of a single tree seed-like collectible in various directions; The information processing device is also used to determine the single feature information of different dimensions corresponding to the single tree seed-type cultural relics based on the single feature image; based on the single feature information, the feature information of each of the tree seed-type cultural relics to be paired, the placement position corresponding to each of the tree seed-type cultural relics to be paired and the pairing model, determine the placement position of the target tree seed-type cultural relics among the tree seed-type cultural relics to be paired that are similar to the single tree seed-type cultural relics in the collection and positioning storage cabinet and the evaluation index of at least one dimension corresponding to the target tree seed-type cultural relics.

4. The automatic pairing system for tree seed-type collectibles according to claim 3 is characterized in that: The acquisition equipment includes: a photography light box, a main frame, a rotating table, a tree seed-like ornament stand, a bottom camera chassis, a top camera and a side camera. The main frame is placed at the bottom of the photography light box, the rotating table is placed at the bottom of the main frame, the tree seed-like ornaments stand is placed on the upper surface of the rotating table, the bottom camera chassis is placed between the rotating table and the tree seed-like ornaments stand, the top camera is placed at the top of the main frame, and the side camera is located on any side of the main frame; The photography light box is used to provide a lighting environment for collecting characteristic images in various directions of the tree seed-type cultural relics to be recorded; the tree seed-type cultural relics to be recorded include one or more of the tree seed-type cultural relics to be paired, the sample tree seed-type cultural relics, and the single tree seed-type cultural relics; The rotating table is used to drive the tree seed-like collectibles to be recorded to rotate; The tree seed type collectibles support is used to place the tree seed type collectibles to be recorded; The bottom camera chassis is used to collect the characteristic image of the tree seed-type collectibles to be recorded in the bottom direction; The top camera is used to collect the characteristic image of the tree seed-type collectibles to be recorded in the top direction; The side camera is used to collect characteristic images of the tree seed-type collectibles to be recorded in various directions on the horizontal plane.

5. The automatic pairing system for tree seed-type collectibles according to any one of claims 1 to 3, characterized in that: The collection and positioning storage cabinet includes: at least one drawer, a cabinet body number, and drawer numbers corresponding to each of the drawers; each of the drawers includes at least one square, and each of the squares includes a row number corresponding to the square and a column number corresponding to the square; the squares correspond one-to-one to the placement positions.

6. The automatic pairing system for tree seed-type collectibles according to any one of claims 1 to 3, characterized in that: Also includes: printer; The printer is connected to the information processing device; The printer is used to print the evaluation index of at least one dimension corresponding to the paired tree seed type cultural relics set and the evaluation index of at least one dimension corresponding to the target tree seed type cultural relics.

7. An automatic pairing method for tree seed-type collectibles, applied to the information processing device in the automatic pairing system for tree seed-type collectibles in claim 1, characterized in that: The method comprises: Acquire characteristic images in all directions of at least one tree seed-type collectible to be paired; Based on each of the characteristic images, characteristic information of each of the tree seed-type cultural relics to be paired in different dimensions is determined; and the placement position of each of the tree seed-type cultural relics to be paired in the collection and positioning storage cabinet is generated; Based on the characteristic information of each of the tree seed-type cultural relics to be paired in different dimensions, the corresponding placement positions and pairing models of each of the tree seed-type cultural relics to be paired, all the tree seed-type cultural relics to be paired are paired, and the target placement position of each paired tree seed-type cultural relic in the paired tree seed-type cultural relic set in the collection and positioning storage cabinet and the evaluation index of at least one dimension corresponding to the paired tree seed-type cultural relic set are determined; the pairing model is obtained by training the initial model of the deep neural network based on the characteristic information of the sample tree seed-type cultural relics; the target placement position is used to instruct the staff to take the paired tree seed-type cultural relic, and the evaluation index is used to characterize the grade of the paired tree seed-type cultural relic and the pricing of the paired tree seed-type cultural relic.

8. The automatic pairing method for tree seed-type collectibles according to claim 7 is characterized in that: The training process of the pairing model includes: Obtaining sample feature images of the sample tree seed-type collectibles in various directions; Based on the sample feature image, determine the sample feature information of the sample tree seed-type collectible and the sample pairing label corresponding to the sample feature information; Inputting the sample feature information into the deep neural network initial model to obtain a model pairing result; Based on the difference between the model pairing result and the sample pairing label, the deep neural network initial model is continuously trained to obtain the pairing model.

9. The automatic pairing method for tree seed-type collectibles according to claim 7 or 8, characterized in that: The method further comprises: Obtain a single feature image of a single tree seed-like collectible in each direction; Based on the single feature image, determining the single feature information of different dimensions corresponding to the single tree seed-type collectible; Based on the single feature information, the feature information of each of the tree seed-type cultural relics to be paired, the placement positions of each of the tree seed-type cultural relics to be paired and the pairing model, the placement position of the target tree seed-type cultural relics among the tree seed-type cultural relics to be paired that are similar to the single tree seed-type cultural relics in the collection and positioning storage cabinet and the evaluation index of at least one dimension corresponding to the target tree seed-type cultural relics are determined.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, it implements the automatic pairing method for tree seed-type collectibles as described in any one of claims 7 to 9.

Images