Item selection method, selection device and terminal equipment

Abstract

The present application is applicable to the field of artificial intelligence technology, and provides an item selection method, a selection device, and a terminal device, the method comprising: obtaining a selection instruction, the selection instruction includes a target item type, obtaining an item image of a candidate item pointed to by the selection instruction, performing type recognition on the candidate item according to the item image, obtaining a candidate item type of the candidate item, screening out preliminary selected items whose corresponding candidate item type belongs to the target item type from the candidate items, performing quality inspection on the preliminary selected items, obtaining quality information of the preliminary selected items, screening out target items whose quality information meets a preset quality standard from the preliminary selected items, and using the target object as the selection result of item selection. The present application can improve the matching degree between the selected items and the user during online shopping.

Description

Technical Field

[0001] The present application belongs to the field of artificial intelligence, and in particular, relates to an item selection method, a selection device and a terminal device. Background Art

[0002] With the development of society, people's pace of life is getting faster and faster. Therefore, in order to reduce shopping time, people usually choose to buy items online. At present, there are generally two ways to shop online: one is to buy items through a web page or application (Application, App), and the other is to send a shopping list to a shopping robot, and the shopping robot will shop according to the shopping list. However, these shopping methods may cause users to buy items that do not meet their needs. For example, when purchasing on a web page or application, it is difficult for users to select items that meet their needs because they cannot see the true appearance of the items. When shopping through a shopping robot, since there are many items of the same type but uneven quality in the market or shopping mall, the shopping robot may also select items that do not meet the user's needs. Therefore, the current online item selection method has the problem of not being able to select items that meet the user's needs, that is, there is a problem of low matching between the selected items and the user. Summary of the invention

[0003] The embodiments of the present application provide an item selection method, a selection device and a terminal device, which can solve the problem in the prior art that the items selected during online shopping have a low matching degree with the user.

[0004] In a first aspect, an embodiment of the present application provides an item selection method, comprising:

[0005] Obtain a selection instruction, wherein the selection instruction includes a target item type;

[0006] Obtaining an object image of a candidate object according to the above selection instruction;

[0007] Performing type recognition on the candidate items according to the item images to obtain candidate item types of the candidate items;

[0008] Selecting preliminary selected items whose corresponding candidate item types belong to the target item type from the candidate items, and performing quality inspection on the preliminary selected items to obtain quality information of the preliminary selected items;

[0009] The target objects whose quality information meets the preset quality standard are selected from the preliminary selected objects, and the target objects are used as the selection results of the object selection.

[0010] In a second aspect, an embodiment of the present application provides an item selection device, comprising:

[0011] A selection instruction acquisition module is used to acquire a selection instruction, wherein the selection instruction includes a target item type;

[0012] An image acquisition module, used to acquire an object image of a candidate object according to the selection instruction;

[0013] an identification module, configured to identify the type of the candidate object according to the object image to obtain the candidate object type of the candidate object;

[0014] A quality information acquisition module, used to select preliminary selected items whose corresponding candidate item types belong to the target item type from the candidate items, and perform quality inspection on the preliminary selected items to obtain quality information of the preliminary selected items;

[0015] The screening module is used to screen out target items whose quality information meets the preset quality standard from the preliminary selected items, and use the target objects as the selection results of item selection.

[0016] In a third aspect, an embodiment of the present application provides a terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in the first aspect when executing the computer program.

[0017] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in the first aspect are implemented.

[0018] In a fifth aspect, an embodiment of the present application provides a computer program product, which, when executed on a terminal device, enables the terminal device to execute the item selection method described in any one of the above-mentioned first aspects.

[0019] In a sixth aspect, an embodiment of the present application provides another method for selecting an item, comprising:

[0020] Obtain a selection instruction, wherein the selection instruction includes a target item type;

[0021] Obtaining an object image of a candidate object according to the above selection instruction;

[0022] Performing type recognition on the candidate items according to the item images to obtain candidate item types of the candidate items;

[0023] Selecting preliminary selected items whose corresponding candidate item types belong to the target item type from the candidate items, and performing quality inspection on the preliminary selected items to obtain quality information of the preliminary selected items;

[0024] Outputting the quality information of the above-mentioned preliminary selected items to the user in a preset output mode;

[0025] A selection instruction from the user based on the quality information feedback is received, a target object pointed to by the selection instruction is selected from the preliminary selected objects, and the target object is used as a selection result of the object selection.

[0026] In a seventh aspect, an embodiment of the present application provides another device for selecting an item, comprising:

[0027] A selection instruction acquisition module is used to acquire a selection instruction, wherein the selection instruction includes a target item type;

[0028] An image acquisition module, used to acquire an object image of a candidate object according to the selection instruction;

[0029] an identification module, configured to identify the type of the candidate object according to the object image to obtain the candidate object type of the candidate object;

[0030] A quality information acquisition module, used to select preliminary selected items whose corresponding candidate item types belong to the target item type from the candidate items, and perform quality inspection on the preliminary selected items to obtain quality information of the preliminary selected items;

[0031] A quality information output module, used to output the quality information of the above-mentioned preliminary selected items to the user in a preset output mode;

[0032] The receiving module is used to receive the selection instruction of the user based on the quality information feedback, select the target object pointed to by the selection instruction from the preliminary selected objects, and use the target object as the selection result of the object selection.

[0033] In an eighth aspect, an embodiment of the present application provides a terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in the sixth aspect when executing the computer program.

[0034] In a ninth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in the sixth aspect are implemented.

[0035] In the tenth aspect, an embodiment of the present application provides a computer program product, which, when executed on a terminal device, enables the terminal device to execute the item selection method described in any one of the sixth aspects above.

[0036] It can be understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here.

[0037] Compared with the prior art, the embodiments of the present application have the following beneficial effects: in the embodiments of the present application, first, a selection instruction is obtained, and the selection instruction contains the target item type. Then, the item image of the candidate item is obtained according to the selection instruction, and the type of the candidate item is identified according to the item image to obtain the candidate item type of the candidate item. The preliminary selected items whose corresponding candidate item type belongs to the target item type are screened out from the candidate items, and the quality of the preliminary selected items is tested to obtain the quality information of the preliminary selected items. Finally, the target items whose quality information meets the preset quality standard are screened out from the preliminary selected items, and the target items are used as the selection result of the item selection. After the preliminary selected items are obtained, the quality of the preliminary selected items is also tested. The preliminary selected items whose quality information meets the preset quality standard are used as the target items. Finally, the target items are used as the result of the item selection. Instead of directly using the preliminary selected items as the result of the item selection. Therefore, effective screening of items can be achieved. At the same time, the preset quality standard can be set according to the needs of the user. Therefore, even if shopping online, the embodiments of the present application can further select items according to the needs of the user, so as to select items with a high degree of matching with the user. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0039] Figure 1 is a flowchart of an item selection method provided in an embodiment of the present application;

[0040] Figure 2 is a schematic diagram of a neural network model provided in one embodiment of the present application;

[0041] Figure 3 is a flowchart of another method for selecting items provided in one embodiment of the present application;

[0042] Figure 4 is a schematic diagram of the structure of an object selection device provided in one embodiment of the present application;

[0043] Figure 5 is a schematic diagram of the structure of another object selection device provided in one embodiment of the present application;

[0044] Figure 6It is a schematic diagram of the structure of the terminal device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0045] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0046] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0047] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0048] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0049] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0050] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0051] The object selection method provided in the embodiment of the present application can be applied to terminal devices such as mobile phones, tablet computers, vehicle-mounted devices, augmented reality (AR) / virtual reality (VR) devices, laptop computers, ultra-mobile personal computers (UMPC), netbooks, personal digital assistants (PDA), etc. The embodiment of the present application does not impose any restrictions on the specific type of the terminal device.

[0052] For example, the terminal device may be a cellular phone, a cordless phone, a Session Initiation Protocol (SIP) phone, a Wireless Local Loop (WLL) station, a Personal Digital Assistant (PDA) device, a handheld device with wireless communication function, a computing device or other processing device connected to a wireless modem, a vehicle-mounted device, a vehicle networking terminal, a computer, a laptop computer, a handheld communication device, a handheld computing device, a satellite wireless device, a TV set top box (STB), a customer premise equipment (CPE) and / or other devices for communicating on a wireless system and a next generation communication system. For example, a mobile terminal in a 5G network or a mobile terminal in a future evolved Public Land Mobile Network (PLMN) network, etc.

[0053] In order to illustrate the technical solution described in this application, a specific embodiment is provided below for illustration.

[0054] Embodiment 1

[0055] The following is a description of an item selection method provided in Example 1 of the present application. Figure 1 , the method comprising:

[0056] Step S101, obtaining a selection instruction, wherein the selection instruction includes a target item type;

[0057] In step S101, the selection instruction includes some requirements for the items that the user needs to purchase. For example, the physical properties of the items and the place of purchase. The specific content of the physical properties of the items is not limited here and can be set according to the actual needs of the user. For example, it can be any one or more of the price, color, quantity and target item type of the items that the user needs to purchase. The above-mentioned target item type refers to the type of item that needs to be selected, for example, apples, grapes and strawberries need to be selected. When the user needs to purchase something, some requirements for the purchased items can be input into the terminal device, and the terminal device generates the corresponding selection instruction.

[0058] Step S102: Acquire an object image of a candidate object according to a selection instruction.

[0059] In step S102, the present application does not specifically limit the source of the item images of the candidate items. The item images of the candidate items may be item images collected by the terminal device camera of the present embodiment at a shopping place after the terminal device of the present embodiment receives a selection instruction, or may be item images collected by a robot at a shopping place after the terminal device of the present embodiment receives a selection instruction, and the robot then sends the collected item images to the terminal device of the present embodiment. It should be noted that the candidate items include target items. For example, when the target item is an apple, the candidate items include apples, pears, oranges, and pitaya. The device terminal of the present embodiment controls the robot to collect item images at a shopping place, and the item images include apple images, pear images, orange images, and pitaya images. After the robot collects the item images, it sends the item images to the device terminal of the present embodiment.

[0060] Step S103: performing type recognition on the candidate item according to the item image to obtain a candidate item type of the candidate item.

[0061] In step S103, when the item image of the candidate item is obtained, it is necessary to identify the type of the candidate item according to the item image to obtain the candidate item type of the candidate item, so as to determine which of the candidate items are the target items. The specific type identification method is not limited here and can be selected or designed by the technician according to actual needs. For example, in some optional embodiments, a type identification method based on a neural network or a type identification method based on a wavelet moment can be used to implement item identification.

[0062] In some possible implementations, the candidate items may be identified by a pre-trained first neural network model, wherein the first neural network model includes a first processing layer, a determiner, and a plurality of second processing layers. First, the first processing layer is used to extract features from the item image to obtain a first feature, and the candidate items are identified by type based on the first feature. If the candidate item type identification fails based on the first feature, a first type set corresponding to the candidate item is identified based on the first feature, and the first type set is input into the determiner, wherein the first type set includes a plurality of item types.

[0063] It should be understood that the above-mentioned first feature may include multiple features of the above-mentioned candidate items. For example, when the candidate item is spinach, the above-mentioned first feature may include green features, leaf features, and root features, etc. The number of the above-mentioned first features can be set according to the features of the items that can be identified by the above-mentioned first processing layer, and the present application does not make specific restrictions here. In addition, the above-mentioned first processing layer and the above-mentioned multiple second processing layers can both identify multiple items at the same time. For example, the above-mentioned first processing layer can identify spinach, water spinach, and Chinese cabbage, etc. The number of items that can be identified by the above-mentioned processing layer can be set according to the actual needs of the user, and the present application does not make specific restrictions here. The first type set corresponding to the above-mentioned candidate items refers to the set of candidate items having the above-mentioned first feature. Although the above-mentioned first processing layer cannot identify what the above-mentioned candidate items are, the above-mentioned first processing layer can obtain the candidate items having the above-mentioned first feature based on the above-mentioned first feature, thereby reducing the amount of calculation for subsequent candidate item type identification, thereby improving the efficiency of type identification.

[0064] For example, when the candidate item is cabbage, the first processing layer extracts features from the cabbage image to obtain the green feature, leaf feature, and flat feature of the cabbage. At this point, the first processing layer can identify that the type of the candidate item is cabbage. If the candidate item is cucumber, the only feature extracted by the first processing layer is the green feature. At this point, based on the green feature, the first processing layer cannot identify the type of the candidate item, i.e., the recognition fails. However, the first processing layer can filter out which items the candidate items may be based on the green feature, i.e., filter out the candidate items with green features, and then input these candidate items with green features into the determiner.

[0065] After receiving the first type set, the determiner extracts features from the item image to obtain a second feature, and then identifies the type of the candidate item based on the second feature and the first type set. If the type identification of the candidate item based on the second feature and the first type set fails, one or more item types are removed from the first type set based on the second feature to obtain a corresponding second type set.

[0066] Among them, the type identification of the candidate items according to the second feature and the first type set includes: type matching of the item types in the first type set according to the second feature, if the match is successful, the successfully matched item type is determined as the candidate item type of the candidate item, if the match fails, it is determined that the type identification of the candidate item fails. For example, when the first type set includes cucumber, winter melon, loofah, bitter melon and pumpkin, if the second feature includes a long strip feature and a thorn feature, the type of the candidate item is matched as cucumber, and the match is successful. If the second feature only includes a long strip, the determiner cannot identify the type of the candidate item at this time, but the determiner can eliminate the item type in the first type set according to the second feature. For example, pumpkin does not have a long strip feature, so pumpkin is eliminated from the first type set to obtain the second type set.

[0067] After obtaining the second type set, the determiner selects a target processing layer corresponding to the second type set from the plurality of second processing layers, and inputs the second type set into the target processing layer. Each second processing layer is used to extract features of one or more candidate items' images and identify the item type corresponding to the candidate items. Selecting the target processing layer corresponding to the second type set from the plurality of second processing layers includes selecting a processing layer that can process each item type in the second type set from the plurality of second processing layers, and using the processing layer that can process each item type in the second type set as the target processing layer.

[0068] Since each processing layer only recognizes preset item types, after the above-mentioned determiner obtains the above-mentioned second type set, it first determines which processing layers in the above-mentioned second processing layers are the target processing layers for identifying the item types in the above-mentioned second type set. After finding the target processing layer corresponding to the above-mentioned second type set, the above-mentioned second type set is directly input into the above-mentioned target processing layer, so that the above-mentioned second type set can be directly input into the above-mentioned target processing layer without passing through other processing layers, thereby improving the efficiency of identifying the types of candidate items.

[0069] Let's take an example to illustrate. Figure 2, assuming that the neural network model has 10 processing layers, but the tenth processing layer is the target processing layer for identifying cucumber images. The existing technical solution is to input the cucumber image to the first processing layer first. If the first processing layer fails to recognize, the cucumber image is input to the second processing layer. If the second processing layer fails to recognize, the cucumber image is input to the third processing layer, the fourth processing layer, and so on until the tenth processing layer. That is, in the existing technical solution, it is not known which processing layer is the target processing layer for the cucumber image. By sequentially passing the cucumber image through the sequentially arranged processing layers (such as Figure 2 201 in the figure), find the target processing layer of the cucumber image. However, this method has a low recognition efficiency. In the present application, the tenth processing layer is determined to be the target processing layer of the cucumber image by the determiner, so that the cucumber image is directly input into the tenth processing layer (as shown in FIG. Figure 2 As shown in 202 in FIG. 1 , the cucumber does not need to go through the processing layer before the tenth processing layer, thereby improving the efficiency of recognition.

[0070] It should be noted that, since each processing layer can be connected to a determiner, when the determiner cannot find the target processing layer of the item type in the second type set, the second type set can be directly input to the next processing layer. If the next processing layer fails to recognize, the second type set is input to the determiner connected to the next processing layer, and the determiner connected to the next processing layer determines which processing layers are the target processing layers for processing the item type in the second type set.

[0071] After obtaining the second type set, the target processing layer extracts features from the object image to obtain a third feature, and identifies the candidate object based on the third feature and the second type set to obtain the candidate object type of the candidate object. Identifying the candidate object based on the third feature and the second type set includes: matching the object types in the second type set based on the third feature, and if the match is successful, determining the matched object type as the object type of the candidate object. If the match fails, it is determined that the identification of the candidate object type fails.

[0072] It should be noted that if the recognition fails, the first neural network model can be jumped out. When the target processing layer is connected to the determiner, when the recognition fails, one or more item types can be removed from the second type set according to the third feature to obtain a third type set, and then the third type set is input to the determiner connected to the target processing layer for further processing.

[0073] The following describes a specific application scenario of the first neural network model:

[0074] When the candidate item is cucumber, the first processing layer extracts features from the cucumber image to obtain a green feature. At this time, according to the green feature, the first processing layer cannot identify the type of the candidate item, that is, the recognition fails, but the first processing layer can identify the candidate item with a green feature. For example, at this time, the candidate items are cucumber, wax gourd, loofah, bitter gourd and pumpkin, and these candidate items with green features are input into the determiner, and the determiner further extracts the second feature. If the second feature is a thorn feature, the type of candidate item corresponding to the cucumber image is determined to be cucumber. If the second feature is a long strip feature, at this time, the determiner cannot determine the type of candidate item corresponding to the cucumber image. However, since the pumpkin does not have a long strip feature, the pumpkin is screened out, and the cucumber, wax gourd, loofah and bitter gourd are input into the target processing layer for processing cucumber, wax gourd, loofah and bitter gourd. The target processing layer further extracts the third feature, and the third feature is a thorn feature, and the type of candidate item corresponding to the cucumber image is determined to be cucumber.

[0075] In this embodiment, the target processing layer for identifying the type of the candidate item is determined by the determiner in the first neural network, and the target processing layer then identifies the object image corresponding to the candidate item, so that the object image corresponding to the candidate item does not need to pass through the sequentially arranged processing layers in turn, thereby improving the efficiency of identifying the object image corresponding to the candidate item.

[0076] Step S104: Screen out preliminary selected items whose corresponding candidate item types belong to the target item type from the candidate items, and perform quality inspection on the preliminary selected items to obtain quality information of the preliminary selected items.

[0077] In step S104, when the candidate item type belongs to the target item type, the candidate items are screened out and used as preliminary selected items. After obtaining the preliminary selected items, the preliminary selected items are subjected to quality inspection to obtain quality information of the preliminary selected items.

[0078] In some embodiments, a pre-trained second neural network model can be used to perform quality inspection on the above-mentioned preliminary selected items, so as to obtain the quality information of the above-mentioned preliminary selected items. The pre-trained second neural network model can adopt models such as Convolutional Neural Network (CNN), Recurrent Neural Networks (RNN) and Deep Neural Network (DNN). The specific type of neural network model used to perform quality inspection on the above-mentioned preliminary selected items can be selected according to actual needs, and this application does not make specific limitations here.

[0079] In other embodiments, the terminal device may first obtain the physical parameters of the above-mentioned preliminary selected items, and then obtain the quality information of the above-mentioned preliminary selected items based on the above-mentioned physical parameters. It should be noted that the above-mentioned physical parameters refer to the non-visual parameters of the above-mentioned items, that is, the parameters that cannot be obtained from the image of the item, including but not limited to the smell, temperature, viscosity, and hardness of the above-mentioned preliminary selected items. It should be understood that when it is necessary to obtain the quality information of the above-mentioned preliminary selected items based on both the visual parameters and the above-mentioned physical parameters, in addition to inputting the above-mentioned item image into the pre-trained second neural network model, it is also necessary to obtain the non-visual parameters of the above-mentioned preliminary selected items in order to obtain the quality information of the above-mentioned preliminary selected items. For example, when the above-mentioned preliminary selected items are tomatoes, in addition to obtaining the price information of tomatoes, it is also necessary to obtain the hardness parameters of tomatoes, and finally obtain the quality information of tomatoes based on the hardness parameters and the price information.

[0080] In some possible implementations, the physical parameters of the preliminary selected items can be obtained by the robot's manipulator. Since at least one sensor is provided on the manipulator, when the manipulator is controlled to touch or grab the preliminary selected items, the sensor will obtain corresponding data, for example, the tactile sensor can obtain the hardness information and viscosity information of the preliminary selected items, and then obtain the quality information of the preliminary selected items based on the data obtained by the sensor.

[0081] Step S105: Filter out target items whose quality information meets the preset quality standard from the preliminarily selected items, and use the target items as the selection results of the item selection.

[0082] In step S105, when the quality information of the preliminarily selected items meets the preset quality standard, the preliminarily selected items are used as target items, and the target items are used as the selection results of item selection.

[0083] In some embodiments, the quality information of the above-mentioned preliminary selected items includes item attribute data. For example, when the above-mentioned preliminary selected items are apples, the item attribute data includes the hardness information, appearance information and price of the apples. The above-mentioned preset quality standard includes multiple attribute standard data set by the user. For example, when the attribute data of the apples include hardness information, appearance information and price, the user can set the attribute standard data of the apples to be hardness A, appearance B, price 6, or the attribute standard data of the apples to be hardness A, appearance B, price 4, etc. The attribute standard data can be set according to the needs of the user, and this application is not specifically limited here. It should be noted that the item attribute data can also include the overall evaluation information of the item. For example, when the item is an apple, the overall evaluation information of the apple is first given according to the hardness information and appearance information of the apple, and then it is judged whether the overall evaluation information and price meet the attribute standard data set by the user. At this time, the user's attribute standard data also needs to include the overall evaluation data of the item. For example, the overall evaluation information of an apple with a hardness of A and an appearance of B is B, and the price is 6. At this time, it is judged whether the apple with an overall evaluation information of B and a price of 6 meets the attribute standard data set by the user.

[0084] Since the user has set multiple attribute standard data, the user can set the priority of each attribute standard data. For example, the attribute standard data with hardness of A, appearance of B, and price of 4 is set as the first priority, the attribute standard data with hardness of A, appearance of B, and price of 6 is set as the second priority, and the attribute standard data with hardness of B, appearance of B, and price of 4 is set as the third priority. Alternatively, the attribute standard data with overall evaluation information of B and price of 4 can be set as the first priority, the attribute standard data with overall evaluation information of B and price of 6 can be set as the second priority, and the attribute standard data with overall evaluation information of A and price of 8 can be set as the third priority. The priority of attribute standard data is set according to the actual needs of the user, and this application does not make any specific restrictions here.

[0085] In the process of screening the above-mentioned preliminary selected items, the attribute data of the above-mentioned items are judged in order from high to low priority to see whether they meet the requirements of the above-mentioned attribute standard data. For example, the first priority is the attribute standard data of hardness A, appearance B, and price 4, and the second priority is the attribute standard data of hardness A, appearance B, and price 6. When the attribute data of the above-mentioned item is hardness A, appearance B, and price 6, first judge whether the attribute data of the above-mentioned item meets the attribute standard data of the first priority. If not, then judge whether the attribute data of the above-mentioned item meets the attribute standard data of the second priority. At this time, if the attribute data of the item meets the attribute standard data of the second priority, then stop judging. If not, continue to judge whether it meets the attribute standard data of other priorities.

[0086] It should be noted that when the user does not set the priority of multiple attribute standard data, in the process of screening the above-mentioned preliminary selected items, the above-mentioned item attribute data needs to be traversed through all attribute standard data until the above-mentioned item attribute data that meets the attribute standard data is found. If the above-mentioned item attribute data does not meet all attribute standard data, a message is sent to the user so that the user can make further selections.

[0087] In other embodiments, user information may be obtained first, and then the preset quality standard may be generated based on the user information, wherein the user information includes one or more of the user's historical item selection records, consumption data, and user identity information. In this case, the preset quality standard is generated based on the user information, and the user does not need to set the preset quality standard, and items with a high degree of matching with the user may be selected, which is more convenient.

[0088] It should be understood that the item selection method in this embodiment can be applied to a terminal device. After the terminal device determines the selection result, it sends the selection result to the robot, and the robot grabs the item according to the selection result. The item selection method in this embodiment can also be applied to a robot. After the robot obtains the selection result itself, it controls the mechanical arm to grab the item according to the selection result.

[0089] As can be seen from the above, the present application provides an item selection method. First, a selection instruction is obtained, and the selection instruction contains a target item type (the target item type refers to the item type to be selected, such as apples, grapes, and strawberries). Then, the item image of the candidate item is obtained according to the selection instruction, and the candidate item is identified according to the item image to obtain the candidate item type of the candidate item (for example, the candidate item type includes apples, pears, and watermelons). The corresponding candidate item type belongs to the target item type from the candidate items (for example, the target item type is apple, and the candidate items include apples, pears, and watermelons, etc., then the apple in the candidate items is determined as the preliminary item), and the preliminary items are quality tested to obtain the quality information of the preliminary items. Finally, the target items whose quality information meets the preset quality standard (the preset quality standard can be set according to user needs) are screened out from the preliminary items, and the above target items are used as the selection result of the item selection. After the preliminary items are obtained, the preliminary items are also quality tested. The preliminary items whose quality information meets the preset quality standard are used as the target items, and finally the target items are used as the result of the item selection. Instead of directly taking the preliminary selected items as the result of item selection. Therefore, effective screening of items can be achieved. At the same time, the preset quality standard can be set according to the needs of the user. Therefore, even if shopping online, the embodiment of the present application can further select items according to the needs of the user, so as to select items with a higher degree of matching with the user.

[0090] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0091] Embodiment 2

[0092] Another method for selecting an item provided in the second embodiment of the present application is described below. Figure 3 , the method comprising:

[0093] S301. Obtain a selection instruction, where the selection instruction includes a target item type.

[0094] S302: Acquire an object image of a candidate object according to a selection instruction.

[0095] S303: Perform type recognition on the object image to obtain candidate item types of the candidate items.

[0096] S304: Filter out preliminary selected items whose corresponding candidate item types belong to the target item type from the candidate items, and perform quality inspection on the preliminary selected items to obtain quality information of the preliminary selected items.

[0097] The specific implementation of steps S301 to S304 is the same as that of steps S101 to S104 in the first embodiment. For details, please refer to the description of the first embodiment, which will not be repeated here.

[0098] S305: Output the quality information of the preliminarily selected items to the user in a preset output mode.

[0099] In step S305, after obtaining the quality information of the above-mentioned preliminary selected items, the above-mentioned quality information is output to the user in a preset output mode so that the user can understand the quality information of the preliminary selected items. It should be noted that, since there are usually many things to be selected, the terminal device can send the quality information of one item to be selected to the user after obtaining it, or it can send the quality information of all items to the user together after obtaining the quality information of all items to be selected. However, when the terminal device obtains the quality information of all items to be selected and then sends the quality information of all items to the user together, after obtaining the quality information of one item, it is necessary to establish a mapping relationship between the quality information of the item and the location information of the item, and store them. So that after obtaining the selection result, the location information of the selected item can be directly obtained, and the selection result and location information are sent to the robot, so that the robot moves to the location of the selected item to grab the item.

[0100] S306: receiving a selection instruction from the user based on the quality information feedback, screening out the target object indicated by the selection instruction from the preliminarily selected objects, and taking the target object as the selection result of the object selection.

[0101] In step S306, after receiving the quality information sent by the terminal device, the user can select items based on the quality information and then feedback the selection instructions to the terminal device. The device terminal then filters out the target items pointed to by the above selection instructions from the last preliminary selected items according to the selection instructions, and uses the above target objects as the selection results of the item selection.

[0102] In some possible implementations, since the location information of the selected item has been mapped to the quality information of the selected item, if the user needs to further interact with the vendor or shopping guide, the user can set the interactive information in the selection instruction. After the device terminal receives the selection instruction, it controls the robot to move to the location of the target item pointed to by the above selection instruction according to the interactive information in the selection instruction. At the same time, the robot can synchronize the user's voice and avatar, so that the user can further interact with the vendor or shopping guide, so that the user can experience remote shopping while selecting items that meet their needs.

[0103] It should be understood that the contents of each optional embodiment in the above-mentioned embodiment 1 can be combined with the present embodiment for application without contradicting the present embodiment. For example, when identifying the type of an item, the first neural network model in the embodiment 1 can be used to identify the type of the item, and the target processing layer of the second type set can be screened out from multiple second processing layers by the determiner, and then the second type set can be directly input into the target processing layer for type identification without passing through other processing layers, thereby improving the efficiency of identifying the candidate item types. For another example, when performing quality inspection on the preliminary selected items, the second neural network model in the embodiment 1 can be used to perform quality inspection on the preliminary selected items, or the quality information of the preliminary selected items can be obtained by obtaining the physical parameters of the preliminary selected items. In addition, the physical parameters of the preliminary selected items can be obtained by the sensors on the robot's manipulator. These combined embodiments should also be included in the scope of protection of the present application.

[0104] In this embodiment, after obtaining the preliminary selected items, the terminal device also performs quality inspection on the preliminary selected items. After obtaining the quality information of the preliminary selected items, the terminal device also outputs the quality information to the user so that the user can select items based on the quality information, and then feeds back the selection instruction to the device terminal. The device terminal selects the target item pointed to by the selection instruction from the preliminary selected items according to the selection instruction, and uses the target object as the selection result of the item selection. The quality information of the preliminary selected items makes it easier for the user to select items that meet their needs.

[0105] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0106] It should be specially explained that, although the first neural network model is used for object type recognition in the above-mentioned embodiments of the present application, the essence of the first neural network model is to recognize the type of objects, wherein the objects can be the objects in the above-mentioned embodiments, or can be human bodies or other animals and plants, etc. Therefore, according to the different needs of actual applications, the first neural network model is applied in combination with more actual scenarios. At this time, it is only necessary to adaptively modify the recognition object of the first neural network model and the selection of the features of the recognition object. At the same time, although the first neural network model has the effective effect of improving the efficiency of object type recognition and making type recognition faster and more accurate compared to the existing object type recognition methods, the beneficial effects of the first neural network model will also undergo appropriate changes under different actual scenarios.

[0107] For example, the first neural network model can be applied to a face recognition scenario. In this case, the object to be recognized is a face. The features used can be one or more of skin color, expression, facial texture, and facial coordinates. Other facial features can also be used, and there are no excessive restrictions here. What is achieved is the classification of faces (such as age, gender, race, or identity classification recognition, etc.). Correspondingly, the beneficial effect of the first neural network model at this time is that it can improve the efficiency of face recognition, making face recognition more accurate and efficient.

[0108] For another example, the first neural network model can also be applied to animal type scenes. In this case, the object to be identified is the animal. The features used can be selected from one or more of the features such as hair, limbs and tail. Other animal features can also be selected, and there is no excessive limitation here. What is achieved is the classification of animals (such as the classification and identification of cats and dogs, or the classification and identification of mammals and reptiles, etc.). Correspondingly, the beneficial effect of the first neural network model at this time is that it can improve the efficiency of animal identification, making animal identification more accurate and efficient.

[0109] In theory, all object type recognition scenarios can be applied in combination with the first neural network model, such as type recognition of different types of vehicles, machines, etc. Therefore, we will not give too many examples here, and the specific ones can be determined according to the actual application scenarios. It should be understood that these different practical applications should all fall within the scope of protection of "using the first neural network model to identify the type of object" in this application.

[0110] Embodiment 3

[0111] Corresponding to the item selection method described in the first embodiment above, Figure 4 The structural block diagram of the object selection device provided in the embodiment of the present application is shown. For the convenience of explanation, only the parts related to the embodiment of the present application are shown. The device 400 includes:

[0112] The selection instruction acquisition module 401 is used to acquire the selection instruction, and the selection instruction includes the target item type.

[0113] The image acquisition module 402 is used to acquire the object image of the candidate object according to the selection instruction.

[0114] The identification module 403 is used to identify the type of the candidate object according to the object image to obtain the candidate object type of the candidate object.

[0115] The quality information acquisition module 404 is used to select preliminary selected items whose corresponding candidate item types belong to the target item type from the candidate items, and perform quality inspection on the preliminary selected items to obtain quality information of the preliminary selected items.

[0116] The screening module 405 is used to screen out target objects whose quality information meets the preset quality standard from the preliminarily selected objects, and use the target objects as the selection results of the object selection.

[0117] Optionally, the identification module 403 is used to execute:

[0118] A pre-trained first neural network model is obtained, wherein the first neural network model includes a first processing layer, a determiner, and a plurality of second processing layers.

[0119] The first processing layer is used to extract features from the object image to obtain a first feature, and the type of the candidate object is identified according to the first feature.

[0120] If the candidate item type is not identified according to the first feature, a first type set corresponding to the candidate item is identified according to the first feature, and the first type set is input to the determiner, where the first type set includes multiple item types.

[0121] The determiner is used to extract features from the object image to obtain a second feature, and the candidate object is identified by type based on the second feature and the first type set.

[0122] If the candidate item type cannot be identified based on the second feature and the first type set, one or more item types are eliminated from the first type set based on the second feature to obtain a corresponding second type set.

[0123] A target processing layer corresponding to the second type set is screened out from the plurality of second processing layers, and the second type set is input into the target processing layer.

[0124] The object processing layer is used to extract features of the object image to obtain a third feature, and the candidate object is identified according to the third feature and the second type set to obtain a candidate object type of the candidate object.

[0125] Optionally, the quality information acquisition module 404 is used to execute:

[0126] The physical parameters of the primary selected objects are obtained.

[0127] The quality information of the initially selected items is obtained according to the physical parameters.

[0128] Optionally, if the item selection device 400 is a robot with a manipulator, and the manipulator is provided with at least one sensor, the quality information acquisition module 404 is used to execute:

[0129] The manipulator is controlled to touch or grasp the preliminary selected object, and the physical parameters are determined according to the data generated by at least one sensor during the touching or grasping process.

[0130] Optionally, the quality information acquisition module 404 is used to execute:

[0131] The object image is input into the pre-trained second neural network model for processing to obtain the quality information of the preliminary selected objects.

[0132] Optionally, the quality information includes item attribute data, and the preset quality standard includes: multiple attribute standard data set by the user. In the screening operation of a single primary selected item, the screening module 405 is used to perform:

[0133] Obtain the priority of each attribute standard data, and determine whether the item attribute data meets the requirements of the attribute standard data in order of priority from high to low.

[0134] If the item attribute data meets the requirements of one or more attribute standard data, the preliminary selected item is determined as the target item.

[0135] Optionally, the screening module 405 includes:

[0136] The user information acquisition unit is used to acquire user information and generate preset quality standards based on the user information. The user information includes one or more of the user's historical item selection records, consumption data, and user identity information.

[0137] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the first embodiment of the method of the present application. Their specific functions and technical effects can be found in the part of the method embodiment and will not be repeated here.

[0138] Embodiment 4

[0139] Corresponding to the item selection method described in the second embodiment above, Figure 5 The structural block diagram of the object selection device provided in the embodiment of the present application is shown. For the convenience of explanation, only the parts related to the embodiment of the present application are shown. The device 500 includes:

[0140] The selection instruction acquisition module 501 is used to acquire the selection instruction, and the selection instruction includes the target item type.

[0141] The image acquisition module 502 is used to acquire the object image of the candidate object according to the selection instruction.

[0142] The identification module 503 is used to identify the type of the candidate object according to the object image to obtain the candidate object type of the candidate object.

[0143] The quality information acquisition module 504 is used to select preliminary selected items whose corresponding candidate item types belong to the target item type from the candidate items, and perform quality inspection on the preliminary selected items to obtain quality information of the preliminary selected items.

[0144] The quality information output module 505 outputs the quality information of the preliminarily selected items to the user in a preset output mode.

[0145] The receiving module 506 is used to receive the selection instruction of the user based on the quality information feedback, select the target object pointed to by the selection instruction from the preliminary selected objects, and use the target object as the selection result of the object selection.

[0146] Optionally, the identification module 503 is used to execute:

[0147] A pre-trained first neural network model is obtained, wherein the first neural network model includes a first processing layer, a determiner, and a plurality of second processing layers.

[0148] The first processing layer is used to extract features from the object image to obtain a first feature, and the type of the candidate object is identified according to the first feature.

[0149] If the candidate item type is not identified according to the first feature, a first type set corresponding to the candidate item is identified according to the first feature, and the first type set is input to the determiner, where the first type set includes multiple item types.

[0150] The determiner is used to extract features from the object image to obtain a second feature, and the candidate object is identified by type based on the second feature and the first type set.

[0151] If the candidate item type cannot be identified based on the second feature and the first type set, one or more item types are eliminated from the first type set based on the second feature to obtain a corresponding second type set.

[0152] A target processing layer corresponding to the second type set is screened out from the plurality of second processing layers, and the second type set is input into the target processing layer.

[0153] The object processing layer is used to extract features of the object image to obtain a third feature, and the candidate object is identified according to the third feature and the second type set to obtain a candidate object type of the candidate object.

[0154] Optionally, the quality information acquisition module 504 is used to execute:

[0155] The physical parameters of the primary selected objects are obtained.

[0156] The quality information of the initially selected items is obtained according to the physical parameters.

[0157] Optionally, if the item selection device 500 is a robot with a manipulator, and the manipulator is provided with at least one sensor, the quality information acquisition module 504 is used to execute:

[0158] The manipulator is controlled to touch or grasp the preliminary selected object, and the physical parameters are determined according to the data generated by at least one sensor during the touching or grasping process.

[0159] Optionally, the quality information acquisition module 504 is used to execute:

[0160] The object image is input into the pre-trained second neural network model for processing to obtain the quality information of the preliminary selected objects.

[0161] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the first embodiment of the method of the present application. Their specific functions and technical effects can be found in the part of the method embodiment and will not be repeated here.

[0162] Embodiment 5

[0163] Figure 6 Schematic diagram of the terminal device provided in Embodiment 5 of the present application. Figure 6 As shown, the terminal device 600 of this embodiment includes: a processor 601, a memory 602, and a computer program 603 stored in the memory 602 and executable on the processor 601. When the processor 601 executes the computer program 603, the steps in the above-mentioned method embodiments are implemented. Alternatively, when the processor 601 executes the computer program 603, the functions of the modules / units in the above-mentioned device embodiments are implemented.

[0164] Exemplarily, the computer program 603 may be divided into one or more modules / units, which are stored in the memory 602 and executed by the processor 601 to complete the present application. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program 603 in the terminal device 600. For example, the computer program 603 may be divided into a selection instruction acquisition module, an image acquisition module, a recognition module, a quality information acquisition module, and a screening module, and the specific functions of each module are as follows:

[0165] Obtaining a selection instruction, wherein the selection instruction includes a target item type;

[0166] Acquire an object image of a candidate object according to the selection instruction;

[0167] Performing type recognition on the candidate object according to the object image to obtain a candidate object type of the candidate object;

[0168] Screening out preliminary selected items whose corresponding candidate item types belong to the target item type from the candidate items, and performing quality inspection on the preliminary selected items to obtain quality information of the preliminary selected items;

[0169] Target objects whose quality information meets the preset quality standard are screened out from the preliminarily selected objects, and the target objects are used as the selection results of the object selection.

[0170] The terminal device may include, but is not limited to, a processor 601 and a memory 602. Those skilled in the art will appreciate that Figure 6 It is only an example of the terminal device 600 and does not constitute a limitation of the terminal device 600. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the terminal device may also include input and output devices, network access devices, buses, etc.

[0171] The processor 601 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware plug-ins, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0172] The memory 602 may be an internal storage unit of the terminal device 600, such as a hard disk or memory of the terminal device 600. The memory 602 may also be an external storage device of the terminal device 600, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the terminal device 600. Further, the memory 602 may also include both an internal storage unit of the terminal device 600 and an external storage device. The memory 602 is used to store the computer program and other programs and data required by the terminal device. The memory 602 may also be used to temporarily store data that has been output or is to be output.

[0173] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the above-mentioned device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0174] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0175] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0176] In the embodiments provided in the present application, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are only schematic. For example, the division of the above modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or plug-ins can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0177] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0178] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0179] If the above-mentioned integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned various method embodiments, and can also be completed by instructing the relevant hardware through a computer program. The above-mentioned computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the above-mentioned computer program includes computer program code, and the above-mentioned computer program code can be in source code form, object code form, executable file or some intermediate form. The above-mentioned computer-readable medium may include: any entity or device capable of carrying the above-mentioned computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the above-mentioned computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0180] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some of the technical features therein by equivalents. 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 application, and should all be included in the protection scope of the present application.

Claims

1. A method for selecting an item, characterized in that: include: Obtaining a selection instruction, wherein the selection instruction includes a target item type; Acquire an object image of a candidate object according to the selection instruction; Performing type recognition on the candidate object according to the object image to obtain a candidate object type of the candidate object; Screening out preliminary selected items whose corresponding candidate item types belong to the target item type from the candidate items, and performing quality inspection on the preliminary selected items to obtain quality information of the preliminary selected items; Screening out target items whose quality information meets a preset quality standard from the preliminarily selected items, and using the target items as selection results of item selection; The step of identifying the type of the candidate item according to the item image to obtain the candidate item type of the candidate item includes: Obtaining a pre-trained first neural network model, wherein the first neural network model includes a first processing layer, a determiner, and a plurality of second processing layers; Using the first processing layer to extract features from the object image to obtain a first feature, and identifying the type of the candidate object according to the first feature; If the candidate item type cannot be identified according to the first feature, a first type set corresponding to the candidate item is identified according to the first feature, and the first type set is input to the determiner, wherein the first type set includes multiple item types; Using the determiner to extract features from the object image to obtain a second feature, and identifying the type of the candidate object based on the second feature and the first type set; If the candidate item type cannot be identified according to the second feature and the first type set, one or more item types are eliminated from the first type set according to the second feature to obtain a corresponding second type set; Filter out a target processing layer corresponding to the second type set from the plurality of second processing layers, and input the second type set to the target processing layer; The target processing layer is used to perform feature extraction on the object image to obtain a third feature, and the candidate object is identified according to the third feature and the second type set to obtain a candidate object type of the candidate object.

2. The method for selecting items according to claim 1, wherein: The step of performing quality inspection on the preliminarily selected items to obtain quality information of the preliminarily selected items includes: Obtaining physical parameters of the primary selected items; The quality information of the initially selected items is obtained according to the physical parameters.

3. The method for selecting items according to claim 2, wherein: Applied to a robot with a manipulator, wherein the manipulator is provided with at least one sensor, the obtaining of the physical parameters of the preliminarily selected object comprises: The manipulator is controlled to touch or grab the preliminarily selected object, and the physical parameter is determined according to the data generated by the at least one sensor during the touching or grabbing process.

4. The method for selecting items according to claim 1, wherein: The step of performing quality inspection on the preliminarily selected items to obtain quality information of the preliminarily selected items includes: The object image is input into a pre-trained second neural network model for processing to obtain the quality information of the primary selected object.

5. The method for selecting items according to claim 1, wherein: The quality information includes item attribute data, and the preset quality standard includes: multiple attribute standard data set by the user; Accordingly, in the operation of selecting the target items whose quality information meets the preset quality standard from the preliminarily selected items, the operation of selecting a single preliminarily selected item includes: Obtaining the priority of each of the attribute standard data, and determining in descending order of priority whether the item attribute data meets the requirements of the attribute standard data; If the item attribute data meets the requirements of one or more of the attribute standard data, the initially selected item is determined as a target item.

6. The method for selecting items according to claim 1, wherein: Also includes: User information is obtained, and the preset quality standard is generated according to the user information, wherein the user information includes one or more of the user's historical item selection records, consumption data, and user identity information.

7. A method for selecting an item, characterized in that: include: Obtaining a selection instruction, wherein the selection instruction includes a target item type; Acquire an object image of a candidate object according to the selection instruction; Performing type recognition on the candidate object according to the object image to obtain a candidate object type of the candidate object; Screening out preliminary selected items whose corresponding candidate item types belong to the target item type from the candidate items, and performing quality inspection on the preliminary selected items to obtain quality information of the preliminary selected items; Outputting the quality information of the preliminarily selected items to the user in a preset output mode; receiving a selection instruction from the user based on the quality information feedback, screening out a target item indicated by the selection instruction from the preliminarily selected items, and using the target item as a selection result of item selection; The step of identifying the type of the candidate item according to the item image to obtain the candidate item type of the candidate item includes: Obtaining a pre-trained first neural network model, wherein the first neural network model includes a first processing layer, a determiner, and a plurality of second processing layers; Using the first processing layer to extract features from the object image to obtain a first feature, and identifying the type of the candidate object according to the first feature; If the candidate item type cannot be identified according to the first feature, a first type set corresponding to the candidate item is identified according to the first feature, and the first type set is input to the determiner, wherein the first type set includes multiple item types; Using the determiner to extract features from the object image to obtain a second feature, and identifying the type of the candidate object based on the second feature and the first type set; If the candidate item type cannot be identified according to the second feature and the first type set, one or more item types are eliminated from the first type set according to the second feature to obtain a corresponding second type set; Filter out a target processing layer corresponding to the second type set from the plurality of second processing layers, and input the second type set to the target processing layer; The target processing layer is used to perform feature extraction on the object image to obtain a third feature, and the candidate object is identified according to the third feature and the second type set to obtain a candidate object type of the candidate object.

8. An object selection device, characterized in that: include: A selection instruction acquisition module, used to acquire a selection instruction, wherein the selection instruction includes a target item type; An image acquisition module, used to acquire an object image of a candidate object according to the selection instruction; an identification module, configured to identify the type of the candidate object according to the object image, and obtain a candidate object type of the candidate object; A quality information acquisition module, used to screen out preliminary selected items whose corresponding candidate item type belongs to the target item type from the candidate items, and perform quality inspection on the preliminary selected items to obtain quality information of the preliminary selected items; A screening module, used to screen out target items whose quality information meets a preset quality standard from the preliminarily selected items, and use the target items as selection results of item selection; The identification module is specifically used for: Obtaining a pre-trained first neural network model, wherein the first neural network model includes a first processing layer, a determiner, and a plurality of second processing layers; Using the first processing layer to extract features from the object image to obtain a first feature, and identifying the type of the candidate object according to the first feature; If the candidate item type cannot be identified according to the first feature, a first type set corresponding to the candidate item is identified according to the first feature, and the first type set is input to the determiner, wherein the first type set includes multiple item types; Using the determiner to extract features from the object image to obtain a second feature, and identifying the type of the candidate object based on the second feature and the first type set; If the candidate item type cannot be identified according to the second feature and the first type set, one or more item types are eliminated from the first type set according to the second feature to obtain a corresponding second type set; Filter out a target processing layer corresponding to the second type set from the plurality of second processing layers, and input the second type set to the target processing layer; The target processing layer is used to perform feature extraction on the object image to obtain a third feature, and the candidate object is identified according to the third feature and the second type set to obtain a candidate object type of the candidate object.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

Images