WO2019096180A1 - 物品识别方法、系统以及电子设备 - Google Patents

物品识别方法、系统以及电子设备 Download PDF

Info

Publication number
WO2019096180A1
WO2019096180A1 PCT/CN2018/115497 CN2018115497W WO2019096180A1 WO 2019096180 A1 WO2019096180 A1 WO 2019096180A1 CN 2018115497 W CN2018115497 W CN 2018115497W WO 2019096180 A1 WO2019096180 A1 WO 2019096180A1
Authority
WO
WIPO (PCT)
Prior art keywords
type
space
probability
spatial
target item
Prior art date
Application number
PCT/CN2018/115497
Other languages
English (en)
French (fr)
Other versions
WO2019096180A9 (zh
Inventor
斯科特·马修·罗伯特
黄鼎隆
傅恺
Original Assignee
深圳码隆科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳码隆科技有限公司 filed Critical 深圳码隆科技有限公司
Publication of WO2019096180A1 publication Critical patent/WO2019096180A1/zh
Publication of WO2019096180A9 publication Critical patent/WO2019096180A9/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/24Classification techniques
    • G06F18/241Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2415Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate

Definitions

  • the present application relates to the field of image recognition technologies, and in particular, to an object identification method, system, and electronic device.
  • Image recognition refers to the technique of using a computer to process, analyze, and understand images to identify targets and objects in various modes.
  • the camera In the process of general image recognition, the camera is used to take pictures, and then the software is used to further identify the processing according to the grayscale difference of the image. Also in geography is a technique for classifying remotely sensed images.
  • the recognition effect of identifying the items in the picture is not accurate enough.
  • some furniture with different functions may be confused due to similar appearance, such as storage of cabinets and bathrooms in the kitchen. Cabinet, which leads to inaccurate recognition results.
  • the purpose of the present application is to provide an item identification method, system, and electronic device, so as to solve the problem that the recognition effect of the current object in the picture is not accurate enough in the prior art, for example, using image recognition technology.
  • image recognition technology When identifying furniture, some furniture with different functions may be confused due to similar appearance, such as kitchen cabinets and bathroom storage cabinets, resulting in inaccurate technical problems in recognition results.
  • an embodiment of the present application provides an item identification method, including:
  • the spatial types include: a first spatial type and a second spatial type;
  • the embodiment of the present application provides the first possible implementation manner of the first aspect, wherein the determining, according to the comparison result, the type of the space to which the target item belongs, obtains an accurate space type, and specifically includes:
  • the comparison result is that the total probability of the first space is greater than the total probability of the second space, determining that the target item belongs to the first space type, and the first space type is an accurate space type;
  • the comparison result is that the total probability of the second space is greater than the total probability of the first space, it is determined that the target item belongs to the second space type, and the second space type is an accurate space type.
  • the embodiment of the present application provides the second possible implementation manner of the first aspect, wherein after the setting the multiple spatial types, the method further includes:
  • the embodiment of the present application provides a third possible implementation manner of the first aspect, wherein the querying the space between the first object remaining space type and the second object remaining space type Type, after getting the shared space type, also includes:
  • the space type to which the target item belongs is confirmed according to the shared space type, and the correct space type is obtained.
  • the embodiment of the present application provides a fourth possible implementation manner of the first aspect, wherein, after determining the type of space to which the target item belongs according to the comparison result, after obtaining the accurate space type, the method further includes:
  • a final space type to which the target item belongs is determined according to the accurate space type and the correct space type.
  • the embodiment of the present application provides the fifth possible implementation manner of the first aspect, wherein after determining the final space type to which the target item belongs according to the accurate space type and the correct space type, include:
  • the item type to which the target item belongs is determined according to the final space type.
  • the embodiment of the present application provides a sixth possible implementation manner of the first aspect, wherein after the setting the multiple spatial types, the method further includes:
  • the embodiment of the present application provides the seventh possible implementation manner of the first aspect, wherein after the identifying and calculating the plane ratio of the object in the image in the space, the method further includes:
  • the spatial type is determined according to the object information and the plane ratio.
  • the embodiment of the present application further provides an item identification system, including:
  • An object recognition module configured to identify an object in an image in which the target item is located, to obtain object information, where the object information includes: first object information and second object information;
  • a space setting module configured to set a plurality of space types, wherein the space type includes: a first space type and a second space type;
  • the probability calculation module is configured to calculate a probability that the first object belongs to the first spatial type and the second spatial type according to the first object information, to obtain a first probability and a second probability;
  • the probability calculation module is further configured to calculate a probability that the second object belongs to the first spatial type and the second spatial type according to the second object information, to obtain a third probability and a fourth probability;
  • a probability statistics module configured to add the first probability to the third probability to obtain a total probability of the first space belonging to the first space type of the target item
  • the probability statistics module is further configured to add the second probability to the fourth probability to obtain a second space total probability that the target item belongs to the second space type;
  • a probability comparison module configured to compare the total probability of the first space with the total probability of the second space to obtain a comparison result
  • a space judging module configured to determine, according to the comparison result, a spatial type to which the target item belongs, to obtain an accurate spatial type
  • An item type judging module configured to determine, according to the accurate space type, an item type to which the target item belongs
  • An initial identification module configured to identify a target item in the image to obtain initial information
  • An item determination module configured to determine accurate information of the target item based on the item type and the initial information.
  • an embodiment of the present application further provides an electronic device, including a memory and a processor, where the computer stores a computer program executable on the processor, where the processor implements the computer program The steps of the method as described above in the first aspect.
  • the item identification method includes: identifying an object in an image of the target item, and obtaining object information;
  • the object information includes: first object information and second object information; setting a plurality of spatial types; wherein the spatial type comprises: a first spatial type and a second spatial type; respectively calculating a first according to the first object information a probability that the object belongs to the first spatial type and the second spatial type, and obtains a first probability and a second probability; respectively, calculating, according to the second object information, that the second object belongs to the first spatial type and the first a probability of the second spatial type, obtaining a third probability and a fourth probability; adding the first probability to the third probability to obtain a total probability of the first space belonging to the first space type of the target item; Adding a second probability to the fourth probability to obtain a second space total probability that the target item belongs to the second space type; Comparing the total
  • Embodiment 1 is a flow chart showing an item identification method provided by Embodiment 1 of the present application.
  • FIG. 2 is a flowchart of an item identification method provided by Embodiment 2 of the present application.
  • FIG. 3 is a schematic structural diagram of an item identification system according to Embodiment 3 of the present application.
  • FIG. 4 is a schematic structural diagram of an electronic device according to Embodiment 4 of the present application.
  • Icons 4-item identification system; 41-object recognition module; 42-space setting module; 43-probability calculation module; 44-probability statistics module; 45-probability comparison module; 46-space judgment module; 47-item type judgment module 48-initial identification module; 49-item determination module; 5-electronic device; 51-memory; 52-processor; 53-bus; 54-communication interface.
  • the recognition effect of identifying the items in the picture is not accurate enough.
  • some furniture with different functions may be confused due to similar appearance, such as kitchen cabinets and bathroom storage cabinets. Therefore, the identification result is inaccurate.
  • an item identification method, system, and electronic device provided by the embodiments of the present application can solve the problem that the recognition of the current item in the prior art is not accurate enough.
  • some furniture with different functions may be confused due to similar appearance, such as kitchen cabinets and bathroom storage cabinets, resulting in inaccurate technical problems of recognition results.
  • Embodiment 1 is a diagrammatic representation of Embodiment 1:
  • An item identification method provided by an embodiment of the present application can be used for identification search of goods such as furniture, and can be applied to scene identification, for example, identifying more accurate items according to scenes.
  • the item identification method may include:
  • S11 Identify an object in an image of the target item to obtain object information; wherein the object information includes: first object information and second object information.
  • the target item may be an item such as furniture that needs to be identified
  • the object in the image may be an object other than the piece of furniture in the image of the piece of furniture, such as an electric appliance, a jewelry, or other furniture.
  • S12 Set a plurality of spatial types; wherein the spatial types include: a first spatial type and a second spatial type.
  • the space type can be a scene type or a room type such as a living room, a dining room, a kitchen or a bedroom.
  • the first space type is a living room and the second space type is a bedroom.
  • S13 Calculate, according to the first object information, a probability that the first object belongs to the first spatial type and the second spatial type, respectively, to obtain a first probability and a second probability.
  • S14 Calculate, according to the second object information, a probability that the second object belongs to the first spatial type and the second spatial type, respectively, to obtain a third probability and a fourth probability.
  • the probability of the scene or room type can be calculated using a scene or room to which the object such as furniture or appliances should belong.
  • the type of scene or room may be living room and bedroom, and the probability of belonging to the living room is calculated to be 50%, and the probability of belonging to the bedroom is 10%.
  • the probability of belonging to the living room is 50%
  • the probability of belonging to the restaurant is 30%
  • the probability of belonging to the kitchen is 10%
  • the probability of belonging to the bedroom is 0%.
  • the total probability that the target item may belong to a certain spatial type is obtained. For example, by calculating the spatial type probability of the TV, the probability of belonging to the living room is 40%, and the probability of belonging to the living room is 50% by calculating the probability of the sofa, and then the respective probabilities belonging to the living room are 40% and 50%. The total probability of getting the space to belong to the living room is 90%.
  • the comparison result may be that the total probability of the first space is greater than the total probability of the second space, or the total probability of the first space may be smaller than the total probability of the second space.
  • S18 Determine the type of space to which the target item belongs according to the comparison result, and obtain an accurate space type.
  • the comparison result is that the total probability of the first space is greater than the total probability of the second space
  • the comparison result is that the total probability of the second space is greater than the total probability of the first space, it is determined that the target item belongs to the second space type, and the second space type is the accurate space type.
  • S21 Determine accurate information of the target item according to the item type and the initial information.
  • Embodiment 2 is a diagrammatic representation of Embodiment 1:
  • An item identification method provided by an embodiment of the present application can be used for identification search of goods such as furniture, and can be applied to scene identification, for example, identifying more accurate items according to scenes.
  • the item identification method may include:
  • S31 Identify an object in an image where the target item is located, and obtain object information.
  • the object information includes: first object information and second object information.
  • the target item may be an item such as furniture that needs to be identified
  • the object in the image may be an object other than the piece of furniture in the image of the piece of furniture, such as an electric appliance, a jewelry, or other furniture.
  • S32 Set multiple spatial types; wherein the spatial types include: a first spatial type and a second spatial type.
  • the space type can be a scene type or a room type such as a living room, a dining room, a kitchen or a bedroom.
  • the first space type is a living room and the second space type is a bedroom.
  • S33 Excluding the spatial type according to the first object information, obtaining a residual space type of the first object; and excluding the spatial type according to the second object information to obtain a remaining space type of the second object.
  • objects such as furniture or appliances that should not be present in each scene or room are used to exclude the type of scene or room.
  • the type of the scene or room excludes the toilet and the bedroom, and the remaining space type of the object may include a living room, a dining room, and a kitchen.
  • S34 Query the same spatial type between the remaining space type of the first object and the remaining spatial type of the second object to obtain a shared spatial type.
  • the remaining space type of the first object includes a living room, a dining room, and a kitchen
  • the remaining space type of the second object includes a living room, a bedroom, and a balcony
  • the shared space type is a living room having both.
  • S35 Confirm the type of space to which the target item belongs according to the shared space type, and obtain the correct space type.
  • the shared space type may be the correct spatial type to which the target item belongs.
  • S36 Calculate, according to the first object information, a probability that the first object belongs to the first spatial type and the second spatial type, and obtain a first probability and a second probability; and calculate, according to the second object information, that the second object belongs to the first spatial type and The probability of the second spatial type yields a third probability and a fourth probability.
  • the probability of the scene or the type of the room is calculated using a scene or room to which the object such as furniture or appliances belongs. For example, if the scene is identified or there is a sofa in the room, the type of scene or room may be living room and bedroom, and the probability of belonging to the living room is calculated to be 50%, and the probability of belonging to the bedroom is 10%. For another example: the identification of the security door is 50% for the living room, 30% for the restaurant, 10% for the kitchen, and 0% for the bedroom.
  • S37 adding the first probability to the third probability to obtain a total probability of the first space belonging to the first space type of the target item; adding the second probability to the fourth probability, to obtain that the target item belongs to the second space type The total probability of the second space.
  • the total probability that the target item may belong to a certain spatial type is obtained by performing an accumulation calculation on the probability of belonging to the same spatial type. For example, by calculating the spatial type probability of the TV, the probability of belonging to the living room is 40%, and the probability of belonging to the living room is 50% by calculating the probability of the sofa, and the respective probabilities belonging to the living room are 40% and 50%. The total probability of getting the space to belong to the living room is 90%.
  • comparison result may be that the total probability of the first space is greater than the total probability of the second space, or that the total probability of the first space is less than the total probability of the second space.
  • S39 Determine the type of space to which the target item belongs according to the comparison result, and obtain an accurate space type.
  • the comparison result is that the total probability of the second space is greater than the total probability of the first space, it is determined that the target item belongs to the second space type, and the second space type is the accurate space type.
  • S40 Determine the final space type to which the target item belongs according to the accurate space type and the correct space type.
  • the exact final spatial type is finally determined.
  • the type of furniture in the figure can be more accurately determined to avoid the confusion of some furniture with different functions. For example, cabinets and bathroom storage cabinets, the final scene can be determined at the end of this step.
  • S41 Determine the item type to which the target item belongs according to the final space type.
  • S43 Determine accurate information of the target item according to the item type and the initial information.
  • the spatial type can be determined by recognizing the percentage of the object in the plane, for example, by recognizing that the proportion of the calculated bed exceeds 60%, and determining the spatial type as the bedroom.
  • Embodiment 3 is a diagrammatic representation of Embodiment 3
  • the item identification system 4 may include: an object identification module 41, a space setting module 42, a probability calculation module 43, a probability and statistics module 44, a probability comparison module 45, The space judging module 46, the item type judging module 47, the initial identifying module 48, and the item determining module 49.
  • the object recognition module 41 it may be configured to identify an object in the image of the target article to obtain object information, wherein the object information includes: first object information and second object information; the space setting module 42 may be configured to set a plurality of spatial types
  • the space type includes: a first space type and a second space type.
  • the probability calculation module 43 may be configured to separately calculate a probability that the first object belongs to the first spatial type and the second spatial type according to the first object information, to obtain a first probability and a second probability; the probability calculation module 43 may also And configured to calculate a probability that the second object belongs to the first spatial type and the second spatial type according to the second object information, to obtain a third probability and a fourth probability.
  • the probability statistics module 44 may be configured to add the first probability to the third probability to obtain a first spatial total probability that the target item belongs to the first spatial type; the probability statistics module 44 may also be configured to be the second The probability is added to the fourth probability to obtain a second spatial total probability that the target item belongs to the second spatial type.
  • the probability comparison module 45 may be configured to compare the first space total probability with the second space total probability to obtain a comparison result; the space determining module 46 may be configured to determine the space type to which the target item belongs according to the comparison result, and obtain The accurate space type; the item type judging module 47 may be configured to determine the item type to which the target item belongs according to the exact space type; the initial identification module 48 may be configured to identify the target item in the image to obtain initial information; the item determining module 49 may be configured to The item type and initial information determine the exact information of the target item.
  • Embodiment 4 is a diagrammatic representation of Embodiment 4:
  • the electronic device 5 may include a memory 51 and a processor 52.
  • the memory 51 stores a computer program executable on the processor 52, and the processor 52 executes the computer.
  • the steps of the method provided in the first embodiment or the second embodiment are implemented in the program.
  • the electronic device 5 may further include a bus 53 and a communication interface 54 connected by a bus 53; the processor 52 is configured to execute an executable module stored in the memory 51, For example, a computer program.
  • the memory 51 may include a high speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory.
  • RAM random access memory
  • non-volatile memory such as at least one disk memory.
  • the communication connection between the system network element and at least one other network element is implemented by at least one communication interface 54 (which may be wired or wireless), and may use an Internet, a wide area network, a local network, a metropolitan area network, or the like.
  • the bus 53 can be an ISA bus, a PCI bus, or an EISA bus.
  • the bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one double-headed arrow is shown in Figure 4, but it does not mean that there is only one bus or one type of bus.
  • the memory 51 may be configured to store a program, and the processor 52 executes the program after receiving the execution instruction, and the method performed by the device defined by the flow process disclosed in any of the foregoing embodiments of the present application may be applied to
  • the processor 52 is implemented by or by the processor 52.
  • Processor 52 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method may be completed by an integrated logic circuit of hardware in the processor 52 or an instruction in the form of software.
  • the processor 52 may be a general-purpose processor, including a central processing unit (CPU) and/or a network processor (NP), etc., or a digital signal processor (Digital Signal Processing, DSP), Application Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic device and/or discrete hardware Component.
  • DSP Central Processing unit
  • ASIC Application Specific Integrated Circuit
  • FPGA Field-Programmable Gate Array
  • the methods, steps, and logical block diagrams disclosed in the embodiments of the present application can be implemented or executed.
  • the general purpose processor may be a microprocessor or the processor or any conventional processor or the like.
  • the steps of the method disclosed in the embodiments of the present application may be directly implemented by the hardware decoding processor, or may be performed by a combination of hardware and software modules in the decoding processor.
  • the software modules can be located in conventional storage media such as random access memory, flash memory, read only memory, programmable read only memory, electrically erasable programmable memory, and/or registers.
  • the storage medium is located in the memory 51, and the processor 52 reads the information in the memory 51 and completes the steps of the above method in combination with its hardware.
  • each block of the flowchart or block diagram can represent a module, a program segment, or a portion of code that comprises one or more of the Executable instructions.
  • the functions noted in the blocks may also occur in a different order than that illustrated in the drawings. For example, two consecutive blocks may be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending upon the functionality involved.
  • each block of the block diagrams and/or flowcharts, and combinations of blocks in the block diagrams and/or flowcharts can be implemented in a dedicated hardware-based system that performs the specified function or function. Or it can be implemented by a combination of dedicated hardware and computer instructions.
  • the electronic device provided by the embodiment of the present application has the same technical features as the item identification method and system provided by the above embodiments, so that the same technical problem can be solved and the same technical effect can be achieved.
  • the terms “installation”, “connected” or “connected” are to be understood broadly, and may be either a fixed connection or a detachable connection, unless explicitly stated and defined otherwise. , or connected integrally; can be mechanical connection or electrical connection; can be directly connected, or can be indirectly connected through an intermediate medium, can be the internal communication of the two elements.
  • installation can be understood in the specific circumstances for those skilled in the art.
  • a computer program product for performing an item identification method comprising a computer readable storage medium storing non-volatile program code executable by a processor, the program code comprising instructions configurable to execute the previous method
  • the disclosed systems, devices, and methods may be implemented in other manners.
  • the device embodiments described above are merely illustrative.
  • the division of the modules is only a logical function division.
  • multiple modules or components may be combined or Can be integrated into another system, or some features can be ignored or not executed.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some communication interface, device or unit, and may be electrical, mechanical or otherwise.
  • the functions may be stored in a computer readable storage medium if implemented in the form of a software functional unit and sold or used as a standalone product.
  • the technical solution of the present application which is essential or contributes to the prior art, or a part of the technical solution, may be embodied in the form of a software product, which is stored in a storage medium, including
  • the instructions are used to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present application.
  • the foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like. .

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Image Analysis (AREA)

Abstract

一种物品识别方法、系统以及电子设备,涉及图像识别技术领域,所述方法包括:识别目标物品所在图像中的物体,得到物体信息(S11);设置多个空间类型(S12);根据第一物体信息分别计算第一物体属于第一空间类型与第二空间类型的概率,得到第一概率与第二概率(S13);根据第二物体信息分别计算第二物体属于第一空间类型与第二空间类型的概率,得到第三概率与第四概率(S14);将第一概率与第三概率进行相加,得到目标物品属于第一空间类型的第一空间总概率(S15);将第二概率与第四概率进行相加,得到目标物品属于第二空间类型的第二空间总概率(S16);将第一空间总概率与第二空间总概率对比,判断目标物品属于的空间类型,判断目标物品属于的物品类型;根据物品类型与初始信息确定目标物品的准确信息(S21),解决了在利用图像识别技术识别家具时部分功能不同的家具会因外形相似而极易造成混淆,从而导致识别结果的不准确的技术问题。

Description

物品识别方法、系统以及电子设备
相关申请的交叉引用
本申请要求于2017年11月14日提交中国专利局的申请号为CN201711126214.3、名称为“物品识别方法、系统以及电子设备”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及图像识别技术领域,尤其是涉及一种物品识别方法、系统以及电子设备。
背景技术
图像识别是指利用计算机对图像进行处理、分析和理解,以识别各种不同模式的目标和对象的技术。
一般图像识别的过程中,采用相机拍摄图片,然后再利用软件根据图片灰阶差做进一步识别处理。另外在地理学中指将遥感图像进行分类的技术。
但是,目前对出图片中的物品进行识别的识别效果不够精准,例如在利用图像识别技术识别家具时,部分功能不同的家具会因外形相似而极易造成混淆,如厨房的橱柜与卫生间的储藏柜,从而导致识别结果的不准确。
发明内容
有鉴于此,本申请的目的在于提供一种物品识别方法、系统以及电子设备,以解决现有技术中存在的目前对出图片中的物品进行识别的识别效果不够精准,例如在利用图像识别技术识别家具时,部分功能不同的家具会因外形相似而极易造成混淆,如厨房的橱柜与卫生间的储藏柜,从而导致识别结果的不准确的技术问题。
第一方面,本申请实施例提供了一种物品识别方法,包括:
识别目标物品所在图像中的物体,得到物体信息;其中,物体信息包括:第一物体信息与第二物体信息;
设置多个空间类型;其中,所述空间类型包括:第一空间类型与第二空间类型;
根据所述第一物体信息分别计算第一物体属于所述第一空间类型与所述第二空间类型的概率,得到第一概率与第二概率;
根据所述第二物体信息分别计算第二物体属于所述第一空间类型与所述第二空间类型的概率,得到第三概率与第四概率;
将所述第一概率与所述第三概率进行相加,得到目标物品属于所述第一空间类型的第一空间总概率;
将所述第二概率与所述第四概率进行相加,得到目标物品属于所述第二空间类型的第 二空间总概率;
将所述第一空间总概率与所述第二空间总概率进行对比,得到对比结果;
根据所述对比结果判断目标物品属于的空间类型,得到准确空间类型;
根据所述准确空间类型判断目标物品属于的物品类型;
识别图像中的目标物品,得到初始信息;
根据所述物品类型与所述初始信息,确定目标物品的准确信息。
结合第一方面,本申请实施例提供了第一方面的第一种可能的实施方式,其中,所述根据所述对比结果判断目标物品属于的空间类型,得到准确空间类型,具体包括:
当所述对比结果为所述第一空间总概率大于所述第二空间总概率时,判断目标物品属于所述第一空间类型,所述第一空间类型为准确空间类型;
当所述对比结果为所述第二空间总概率大于所述第一空间总概率时,判断目标物品属于所述第二空间类型,所述第二空间类型为准确空间类型。
结合第一方面,本申请实施例提供了第一方面的第二种可能的实施方式,其中,所述设置多个空间类型之后,还包括:
根据所述第一物体信息对所述空间类型进行排除,得到第一物体剩余空间类型;
根据所述第二物体信息对所述空间类型进行排除,得到第二物体剩余空间类型;
查询所述第一物体剩余空间类型与所述第二物体剩余空间类型之间相同的空间类型,得到共有空间类型。
结合第一方面,本申请实施例提供了第一方面的第三种可能的实施方式,其中,所述查询所述第一物体剩余空间类型与所述第二物体剩余空间类型之间相同的空间类型,得到共有空间类型之后,还包括:
根据所述共有空间类型确认目标物品属于的空间类型,得到正确空间类型。
结合第一方面,本申请实施例提供了第一方面的第四种可能的实施方式,其中,所述根据所述对比结果判断目标物品属于的空间类型,得到准确空间类型之后,还包括:
根据所述准确空间类型与所述正确空间类型判断目标物品属于的最终空间类型。
结合第一方面,本申请实施例提供了第一方面的第五种可能的实施方式,其中,所述根据所述准确空间类型与所述正确空间类型判断目标物品属于的最终空间类型之后,还包括:
根据所述最终空间类型判断目标物品属于的物品类型。
结合第一方面,本申请实施例提供了第一方面的第六种可能的实施方式,其中,所述设置多个空间类型之后,还包括:
识别与计算图像中的物体在空间中所占的平面比例。
结合第一方面,本申请实施例提供了第一方面的第七种可能的实施方式,其中,所述识别与计算图像中的物体在空间中所占的平面比例之后,还包括:
根据所述物体信息与所述平面比例判断空间类型。
第二方面,本申请实施例还提供一种物品识别系统,包括:
物体识别模块,配置成识别目标物品所在图像中的物体,得到物体信息,其中,物体信息包括:第一物体信息与第二物体信息;
空间设置模块,配置成设置多个空间类型,其中,所述空间类型包括:第一空间类型与第二空间类型;
概率计算模块,配置成根据所述第一物体信息分别计算第一物体属于所述第一空间类型与所述第二空间类型的概率,得到第一概率与第二概率;
概率计算模块还配置成根据所述第二物体信息分别计算第二物体属于所述第一空间类型与所述第二空间类型的概率,得到第三概率与第四概率;
概率统计模块,配置成将所述第一概率与所述第三概率进行相加,得到目标物品属于所述第一空间类型的第一空间总概率;
概率统计模块还配置成将所述第二概率与所述第四概率进行相加,得到目标物品属于所述第二空间类型的第二空间总概率;
概率对比模块,配置成将所述第一空间总概率与所述第二空间总概率进行对比,得到对比结果;
空间判断模块,配置成根据所述对比结果判断目标物品属于的空间类型,得到准确空间类型;
物品类型判断模块,配置成根据所述准确空间类型判断目标物品属于的物品类型;
初始识别模块,配置成识别图像中的目标物品,得到初始信息;
物品确定模块,配置成根据所述物品类型与所述初始信息,确定目标物品的准确信息。
第三方面,本申请实施例还提供一种电子设备,包括存储器和处理器,所述存储器中存储有可在所述处理器上运行的计算机程序,所述处理器执行所述计算机程序时实现上述如第一方面所述的方法的步骤。
本申请实施例提供的技术方案带来了以下有益效果:本申请实施例提供的物品识别方法、系统以及电子设备中,物品识别方法包括:识别目标物品所在图像中的物体,得到物体信息;其中,物体信息包括:第一物体信息与第二物体信息;设置多个空间类型;其中,所述空间类型包括:第一空间类型与第二空间类型;根据所述第一物体信息分别计算第一物体属于所述第一空间类型与所述第二空间类型的概率,得到第一概率与第二概率;根据所述第二物体信息分别计算第二物体属于所述第一空间类型与所述第二空间类型的概率, 得到第三概率与第四概率;将所述第一概率与所述第三概率进行相加,得到目标物品属于所述第一空间类型的第一空间总概率;将所述第二概率与所述第四概率进行相加,得到目标物品属于所述第二空间类型的第二空间总概率;将所述第一空间总概率与所述第二空间总概率进行对比,得到对比结果;根据所述对比结果判断目标物品属于的空间类型,得到准确空间类型;根据所述准确空间类型判断目标物品属于的物品类型;识别图像中的目标物品,得到初始信息;根据所述物品类型与所述初始信息,确定目标物品的准确信息,在图像识别过程中,通过对与目标物品相同空间的其他物体也进行识别,从而判断出该空间的类型,以此来辨别出目标识别物品的类型与功能,再根据该类型以及对目标物品初步识别出的外形等,得到更加精确的目标物品识别结果,在此过程中,通过对目标物品的类型的判断,使得功能不同的物品不会再因外形相似而造成混淆,从而解决了现有技术中存在的目前对出图片中的物品进行识别的识别效果不够精准,例如在利用图像识别技术识别家具时,部分功能不同的家具会因外形相似而极易造成混淆,如厨房的橱柜与卫生间的储藏柜,从而导致识别结果的不准确的技术问题。
本申请的其他特征和优点将在随后的说明书中阐述,并且,部分地从说明书中变得显而易见,或者通过实施本申请而了解。本申请的目的和其他优点在说明书、权利要求书以及附图中所特别指出的结构来实现和获得。
为使本申请的上述目的、特征和优点能更明显易懂,下文特举较佳实施例,并配合所附附图,作详细说明如下。
附图说明
为了更清楚地说明本申请具体实施方式或现有技术中的技术方案,下面将对具体实施方式或现有技术描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图是本申请的一些实施方式,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1示出了本申请实施例一所提供的物品识别方法的流程图;
图2示出了本申请实施例二所提供的物品识别方法的流程图;
图3示出了本申请实施例三所提供的一种物品识别系统的结构示意图;
图4示出了本申请实施例四所提供的一种电子设备的结构示意图。
图标:4-物品识别系统;41-物体识别模块;42-空间设置模块;43-概率计算模块;44-概率统计模块;45-概率对比模块;46-空间判断模块;47-物品类型判断模块;48-初始识别模块;49-物品确定模块;5-电子设备;51-存储器;52-处理器;53-总线;54-通信接口。
具体实施方式
为使本申请实施例的目的、技术方案和优点更加清楚,下面将结合附图对本申请的技术方案进行清楚且完整地描述,显然,所描述的实施例是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
目前对出图片中的物品进行识别的识别效果不够精准,例如在利用图像识别技术识别家具时,部分功能不同的家具会因外形相似而极易造成混淆,如厨房的橱柜与卫生间的储藏柜,从而导致识别结果的不准确,基于此,本申请实施例提供的一种物品识别方法、系统以及电子设备,可以解决现有技术中存在的目前对出图片中的物品进行识别的识别效果不够精准,例如在利用图像识别技术识别家具时,部分功能不同的家具会因外形相似而极易造成混淆,如厨房的橱柜与卫生间的储藏柜,从而导致识别结果的不准确的技术问题。
为便于对本实施例进行理解,首先对本申请实施例所公开的一种物品识别方法、系统以及电子设备进行详细介绍。
实施例一:
本申请实施例提供的一种物品识别方法,可以用于家具等商品的识别搜索,能够应用于场景的识别,例如根据场景识别出更精准的商品。如图1所示,物品识别方法可以包括:
S11:识别目标物品所在图像中的物体,得到物体信息;其中,物体信息包括:第一物体信息与第二物体信息。
作为一个优选方案,识别目标物品所在图像中的除目标物品以外的其他物体。目标物品可以为需要识别的家具等商品,图像中的物体可以为该家具所在图像中的除该家具以外的其他物体,如电器、摆饰品或其他家具等。
S12:设置多个空间类型;其中,空间类型包括:第一空间类型与第二空间类型。
其中,空间类型可以为客厅、餐厅、厨房或卧室等场景类型或房间类型。例如,第一空间类型为客厅,第二空间类型为卧室。
S13:根据第一物体信息分别计算第一物体属于第一空间类型与第二空间类型的概率,得到第一概率与第二概率。
S14:根据第二物体信息分别计算第二物体属于第一空间类型与第二空间类型的概率,得到第三概率与第四概率。
优选的,可以利用家具或电器等物体应该属于的场景或房间,来计算场景或房间类型的概率。例如,如果识别到该场景或房间内有沙发,则该场景或房间的类型可能会客厅与卧室,经过计算得出属于客厅的概率为50%,属于卧室的概率为10%。再例如:识别到防 盗门,属于客厅的概率为50%,属于餐厅的概率为30%,属于厨房的概率为10%,属于卧室的概率为0%。
S15:将第一概率与第三概率进行相加,得到目标物品属于第一空间类型的第一空间总概率。
S16:将第二概率与第四概率进行相加,得到目标物品属于第二空间类型的第二空间总概率。
具体的,通过对属于相同空间类型之间概率的累加计算,得到目标物品可能属于某个空间类型的总概率。例如,通过对电视的空间类型概率计算得到属于客厅的概率为40%,通过对沙发的概率计算得到属于客厅的概率为50%,则将同属于客厅的各个概率40%与50%进行相加,得到该空间属于客厅的总概率为90%。
S17:将第一空间总概率与第二空间总概率进行对比,得到对比结果。
本步骤中,对比结果可以为第一空间总概率大于第二空间总概率,也可以为第一空间总概率小于第二空间总概率。
S18:根据对比结果判断目标物品属于的空间类型,得到准确空间类型。
作为本实施例的优选实施方式,当对比结果为第一空间总概率大于第二空间总概率时,判断目标物品属于第一空间类型,第一空间类型为准确空间类型。
同样的,当对比结果为第二空间总概率大于第一空间总概率时,判断目标物品属于第二空间类型,第二空间类型为准确空间类型。
S19:根据准确空间类型判断目标物品属于的物品类型。
S20:识别图像中的目标物品,得到初始信息。
S21:根据物品类型与初始信息,确定目标物品的准确信息。
在实际应用中,能够根据每个场景应该有什么,通过各种筛选确定与概率计算,最终确定需要识别的物品所属空间类型,从而便于确定后续的目标物品的准确信息。
因此,在识别出图片中的场景属于何种室内空间后,如客厅、卧室、卫生间或厨房等,能够根据所在空间能够更精准的判断出图中家具的类型而避免部分功能不同的家具因外形相似而混淆,例如橱柜与卫生间储藏柜,虽然外形都属于柜子,但由于放置于不通过的房间而具有不同的具体功能,通过该方法便能够精确的识别出该柜子为厨房的橱柜还是卫生间的储藏柜。
实施例二:
本申请实施例提供的一种物品识别方法,可以用于家具等商品的识别搜索,能够应用于场景的识别,例如根据场景识别出更精准的商品。如图2所示,物品识别方法可以包括:
S31:识别目标物品所在图像中的物体,得到物体信息;其中,物体信息包括:第一物体信息与第二物体信息。
本步骤中,识别目标物品所在图像中的除目标物品以外的其他物体。目标物品可以为需要识别的家具等商品,图像中的物体可以为该家具所在图像中的除该家具以外的其他物体,如电器、摆饰品或其他家具等。
S32:设置多个空间类型;其中,空间类型包括:第一空间类型与第二空间类型。
其中,空间类型可以为客厅、餐厅、厨房或卧室等场景类型或房间类型。例如,第一空间类型为客厅,第二空间类型为卧室。
S33:根据第一物体信息对空间类型进行排除,得到第一物体剩余空间类型;根据第二物体信息对空间类型进行排除,得到第二物体剩余空间类型。
本实施例中,利用每个场景或房间不应该有的家具或电器等物体来排除场景或房间的类型。例如,如果该场景或房间内有冰箱,则该场景或房间的类型排除厕所与卧室,物体剩余空间类型便可以包括客厅、餐厅与厨房。
S34:查询第一物体剩余空间类型与第二物体剩余空间类型之间相同的空间类型,得到共有空间类型。
例如,经过类型排除后,第一物体剩余空间类型包括客厅、餐厅与厨房,第二物体剩余空间类型包括客厅、卧室与阳台,则共有空间类型便为二者都具有的客厅。
S35:根据共有空间类型确认目标物品属于的空间类型,得到正确空间类型。
作为本实施例的优选实施方式,共有空间类型可以为目标物品属于的正确空间类型。
S36:根据第一物体信息分别计算第一物体属于第一空间类型与第二空间类型的概率,得到第一概率与第二概率;根据第二物体信息分别计算第二物体属于第一空间类型与第二空间类型的概率,得到第三概率与第四概率。
进一步的是,利用家具或电器等物体应该属于的场景或房间,来计算场景或房间类型的概率。例如,如果识别到该场景或房间内有沙发,则该场景或房间的类型可能会客厅与卧室,经过计算得出属于客厅的概率为50%,属于卧室的概率为10%。再例如:识别到防盗门,属于客厅的概率为50%,属于餐厅的概率为30%,属于厨房的概率为10%,属于卧室的概率为0%。
S37:将第一概率与第三概率进行相加,得到目标物品属于第一空间类型的第一空间总概率;将第二概率与第四概率进行相加,得到目标物品属于第二空间类型的第二空间总概率。
优选的,通过对属于相同空间类型之间概率的累加计算,得到目标物品可能属于某个空间类型的总概率。例如,通过对电视的空间类型概率计算得到属于客厅的概率为40%, 通过对沙发的概率计算得到属于客厅的概率为50%,则将同属于客厅的各个概率40%与50%进行相加,得到该空间属于客厅的总概率为90%。
S38:将第一空间总概率与第二空间总概率进行对比,得到对比结果。
需要说明的是,对比结果可以为第一空间总概率大于第二空间总概率,也可以为第一空间总概率小于第二空间总概率。
S39:根据对比结果判断目标物品属于的空间类型,得到准确空间类型。
在一种实现方式中,当对比结果为第一空间总概率大于第二空间总概率时,判断目标物品属于第一空间类型,第一空间类型为准确空间类型。
在另一种实现方式中,当对比结果为第二空间总概率大于第一空间总概率时,判断目标物品属于第二空间类型,第二空间类型为准确空间类型。
S40:根据准确空间类型与正确空间类型判断目标物品属于的最终空间类型。
进一步,根据通过步骤S39得到的准确空间类型与通过步骤S35得到的正确空间类型,最终确定精确的最终空间类型。在识别出图片中的场景属于何种室内空间后,如客厅、卧室、卫生间或厨房等,根据所在空间能够更精准的判断出图中家具的类型而避免部分功能不同的家具因外形相似而混淆,例如橱柜与卫生间储藏柜,这步的最后便能够确定了最终场景。
S41:根据最终空间类型判断目标物品属于的物品类型。
S42:识别图像中的目标物品,得到初始信息。
S43:根据物品类型与初始信息,确定目标物品的准确信息。
本实施例中,能够根据每个场景应该有什么与不应该有什么,通过各种排除与确定,还可以利用概率,最终确定需要识别的物品所属空间类型,从而便于确定后续的目标物品的准确信息。
作为本实施例的另一种实施方式,在设置多个空间类型之后,识别与计算图像中的物体在空间中所占的平面比例,然后,根据物体信息与平面比例判断空间类型。因此,可以通过识别计算该物体在平面所占的百分比来判断空间类型,例如识别计算出床所占得比例超过60%,则判断该空间类型为卧室。
实施例三:
本申请实施例提供的一种物品识别系统,如图3所示,物品识别系统4可以包括:物体识别模块41、空间设置模块42、概率计算模块43、概率统计模块44、概率对比模块45、空间判断模块46、物品类型判断模块47、初始识别模块48以及物品确定模块49。
对于物体识别模块41,可以配置成识别目标物品所在图像中的物体,得到物体信息, 其中,物体信息包括:第一物体信息与第二物体信息;空间设置模块42可以配置成设置多个空间类型,其中,空间类型包括:第一空间类型与第二空间类型。
在实际应用中,概率计算模块43可以配置成根据第一物体信息分别计算第一物体属于第一空间类型与第二空间类型的概率,得到第一概率与第二概率;概率计算模块43还可以配置成根据第二物体信息分别计算第二物体属于第一空间类型与第二空间类型的概率,得到第三概率与第四概率。
作为一个优选方案,概率统计模块44可以配置成将第一概率与第三概率进行相加,得到目标物品属于第一空间类型的第一空间总概率;概率统计模块44还可以配置成将第二概率与第四概率进行相加,得到目标物品属于第二空间类型的第二空间总概率。
需要说明的是,概率对比模块45可以配置成将第一空间总概率与第二空间总概率进行对比,得到对比结果;空间判断模块46可以配置成根据对比结果判断目标物品属于的空间类型,得到准确空间类型;物品类型判断模块47可以配置成根据准确空间类型判断目标物品属于的物品类型;初始识别模块48可以配置成识别图像中的目标物品,得到初始信息;物品确定模块49可以配置成根据物品类型与初始信息,确定目标物品的准确信息。
实施例四:
本申请实施例提供的一种电子设备,如图4所示,电子设备5可以包括存储器51和处理器52,存储器51中存储有可在处理器52上运行的计算机程序,处理器52执行计算机程序时实现上述实施例一或实施例二提供的方法的步骤。
参见图4,电子设备5还可以包括:总线53和通信接口54,所述处理器52、通信接口54和存储器51通过总线53连接;处理器52配置成执行存储器51中存储的可执行模块,例如计算机程序。
其中,存储器51可能包含高速随机存取存储器(RAM,Random Access Memory),也可能还包括非不稳定的存储器(non-volatile memory),例如至少一个磁盘存储器。通过至少一个通信接口54(可以是有线或者无线)实现该系统网元与至少一个其他网元之间的通信连接,可以使用互联网,广域网,本地网,城域网等。
总线53可以是ISA总线、PCI总线或EISA总线等。所述总线可以分为地址总线、数据总线和控制总线等。为便于表示,图4中仅用一个双向箭头表示,但并不表示仅有一根总线或一种类型的总线。
其中,存储器51可以配置成存储程序,所述处理器52在接收到执行指令后,执行所述程序,前述本申请实施例任一实施例揭示的流过程定义的装置所执行的方法可以应用于处理器52中,或者由处理器52实现。
处理器52可能是一种集成电路芯片,具有信号的处理能力。在实现过程中,上述方法的各步骤可以通过处理器52中的硬件的集成逻辑电路或者软件形式的指令完成。上述的处理器52可以是通用处理器,包括中央处理器(Central Processing Unit,简称CPU)和/或网络处理器(Network Processor,简称NP)等;还可以是数字信号处理器(Digital Signal Processing,简称DSP)、专用集成电路(Application Specific Integrated Circuit,简称ASIC)、现成可编程门阵列(Field-Programmable Gate Array,简称FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件和/或分立硬件组件。可以实现或者执行本申请实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。结合本申请实施例所公开的方法的步骤可以直接体现为硬件译码处理器执行完成,或者用译码处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存储器、闪存、只读存储器、可编程只读存储器、电可擦写可编程存储器和/或寄存器等本领域成熟的存储介质中。该存储介质位于存储器51,处理器52读取存储器51中的信息,结合其硬件完成上述方法的步骤。
除非另外具体说明,否则在这些实施例中阐述的部件和步骤的相对步骤、数字表达式和数值并不限制本申请的范围。
本申请实施例所提供的装置,其实现原理及产生的技术效果和前述方法实施例相同,为简要描述,装置实施例部分未提及之处,可参考前述方法实施例中相应内容。
在这里示出和描述的所有示例中,任何具体值应被解释为仅仅是示例性的,而不是作为限制,因此,示例性实施例的其他示例可以具有不同的值。
附图中的流程图和框图显示了根据本申请的多个实施例的系统、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上,流程图或框图中的每个方框可以代表一个模块、程序段或代码的一部分,所述模块、程序段或代码的一部分包含一个或多个配置成实现规定的逻辑功能的可执行指令。也应当注意,在有些作为替换的实现中,方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如,两个连续的方框实际上可以基本并行地执行,它们有时也可以按相反的顺序执行,这依所涉及的功能而定。也要注意的是,框图和/或流程图中的每个方框、以及框图和/或流程图中的方框的组合,可以用执行规定的功能或动作的专用的基于硬件的系统来实现,或者可以用专用硬件与计算机指令的组合来实现。
本申请实施例提供的电子设备,与上述实施例提供的物品识别方法、以及系统具有相同的技术特征,所以也能解决相同的技术问题,达到相同的技术效果。
另外,在本申请实施例的描述中,除非另有明确的规定和限定,术语“安装”、“相连”或“连接”应做广义理解,例如,可以是固定连接,也可以是可拆卸连接,或一体地连接; 可以是机械连接,也可以是电连接;可以是直接相连,也可以通过中间媒介间接相连,可以是两个元件内部的连通。对于本领域的普通技术人员而言,可以具体情况理解上述术语在本申请中的具体含义。
此外,术语“第一”、“第二”或“第三”仅用于描述目的,而不能理解为指示或暗示相对重要性。
本申请实施例所提供的进行物品识别方法的计算机程序产品,包括存储了处理器可执行的非易失的程序代码的计算机可读存储介质,所述程序代码包括的指令可配置成执行前面方法实施例中所述的方法,具体实现可参见方法实施例,在此不再赘述。
在本申请所提供的几个实施例中,应该理解到,所揭露的系统、装置和方法,可以通过其它的方式实现。以上所描述的装置实施例仅仅是示意性的,例如,所述模块的划分,仅仅为一种逻辑功能划分,实际实现时可以有另外的划分方式,又例如,多个模块或组件可以结合或者可以集成到另一个系统,或一些特征可以忽略,或不执行。另一点,所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些通信接口,装置或单元的间接耦合或通信连接,可以是电性,机械或其它的形式。
所述功能如果以软件功能单元的形式实现并作为独立的产品销售或使用时,可以存储在一个计算机可读取存储介质中。基于这样的理解,本申请的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的部分可以以软件产品的形式体现出来,该计算机软件产品存储在一个存储介质中,包括若干指令用以使得一台计算机设备(可以是个人计算机,服务器,或者网络设备等)执行本申请各个实施例所述方法的全部或部分步骤。而前述的存储介质包括:U盘、移动硬盘、只读存储器(ROM,Read-Only Memory)、随机存取存储器(RAM,Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
最后应说明的是:以上所述实施例,仅为本申请的具体实施方式,用以说明本申请的技术方案,而非对其限制,本申请的保护范围并不局限于此,尽管参照前述实施例对本申请进行了详细的说明,本领域的普通技术人员应当理解:任何熟悉本技术领域的技术人员在本申请揭露的技术范围内,其依然可以对前述实施例所记载的技术方案进行修改或可轻易想到变化,或者对其中部分技术特征进行等同替换;而这些修改、变化或者替换,并不使相应技术方案的本质脱离本申请实施例技术方案的精神和范围,都应涵盖在本申请的保护范围之内。因此,本申请的保护范围应所述以权利要求的保护范围为准。

Claims (10)

  1. 一种物品识别方法,其特征在于,包括:
    识别目标物品所在图像中的物体,得到物体信息;其中,物体信息包括:第一物体信息与第二物体信息;
    设置多个空间类型;其中,所述空间类型包括:第一空间类型与第二空间类型;
    根据所述第一物体信息分别计算第一物体属于所述第一空间类型与所述第二空间类型的概率,得到第一概率与第二概率;
    根据所述第二物体信息分别计算第二物体属于所述第一空间类型与所述第二空间类型的概率,得到第三概率与第四概率;
    将所述第一概率与所述第三概率进行相加,得到目标物品属于所述第一空间类型的第一空间总概率;
    将所述第二概率与所述第四概率进行相加,得到目标物品属于所述第二空间类型的第二空间总概率;
    将所述第一空间总概率与所述第二空间总概率进行对比,得到对比结果;
    根据所述对比结果判断目标物品属于的空间类型,得到准确空间类型;
    根据所述准确空间类型判断目标物品属于的物品类型;
    识别图像中的目标物品,得到初始信息;
    根据所述物品类型与所述初始信息,确定目标物品的准确信息。
  2. 根据权利要求1所述的物品识别方法,其特征在于,所述根据所述对比结果判断目标物品属于的空间类型,得到准确空间类型,具体包括:
    当所述对比结果为所述第一空间总概率大于所述第二空间总概率时,判断目标物品属于所述第一空间类型,所述第一空间类型为准确空间类型;
    当所述对比结果为所述第二空间总概率大于所述第一空间总概率时,判断目标物品属于所述第二空间类型,所述第二空间类型为准确空间类型。
  3. 根据权利要求1所述的物品识别方法,其特征在于,所述设置多个空间类型之后,还包括:
    根据所述第一物体信息对所述空间类型进行排除,得到第一物体剩余空间类型;
    根据所述第二物体信息对所述空间类型进行排除,得到第二物体剩余空间类型;
    查询所述第一物体剩余空间类型与所述第二物体剩余空间类型之间相同的空间类型,得到共有空间类型。
  4. 根据权利要求3所述的物品识别方法,其特征在于,所述查询所述第一物体剩 余空间类型与所述第二物体剩余空间类型之间相同的空间类型,得到共有空间类型之后,还包括:
    根据所述共有空间类型确认目标物品属于的空间类型,得到正确空间类型。
  5. 根据权利要求4所述的物品识别方法,其特征在于,所述根据所述对比结果判断目标物品属于的空间类型,得到准确空间类型之后,还包括:
    根据所述准确空间类型与所述正确空间类型判断目标物品属于的最终空间类型。
  6. 根据权利要求5所述的物品识别方法,其特征在于,所述根据所述准确空间类型与所述正确空间类型判断目标物品属于的最终空间类型之后,还包括:
    根据所述最终空间类型判断目标物品属于的物品类型。
  7. 根据权利要求1所述的物品识别方法,其特征在于,所述设置多个空间类型之后,还包括:
    识别与计算图像中的物体在空间中所占的平面比例。
  8. 根据权利要求7所述的物品识别方法,其特征在于,所述识别与计算图像中的物体在空间中所占的平面比例之后,还包括:
    根据所述物体信息与所述平面比例判断空间类型。
  9. 一种物品识别系统,其特征在于,包括:
    物体识别模块,配置成识别目标物品所在图像中的物体,得到物体信息,其中,物体信息包括:第一物体信息与第二物体信息;
    空间设置模块,配置成设置多个空间类型,其中,所述空间类型包括:第一空间类型与第二空间类型;
    概率计算模块,配置成根据所述第一物体信息分别计算第一物体属于所述第一空间类型与所述第二空间类型的概率,得到第一概率与第二概率;
    概率计算模块还配置成根据所述第二物体信息分别计算第二物体属于所述第一空间类型与所述第二空间类型的概率,得到第三概率与第四概率;
    概率统计模块,配置成将所述第一概率与所述第三概率进行相加,得到目标物品属于所述第一空间类型的第一空间总概率;
    概率统计模块还配置成将所述第二概率与所述第四概率进行相加,得到目标物品属于所述第二空间类型的第二空间总概率;
    概率对比模块,配置成将所述第一空间总概率与所述第二空间总概率进行对比,得到对比结果;
    空间判断模块,配置成根据所述对比结果判断目标物品属于的空间类型,得到准确空间类型;
    物品类型判断模块,配置成根据所述准确空间类型判断目标物品属于的物品类型;
    初始识别模块,配置成识别图像中的目标物品,得到初始信息;
    物品确定模块,配置成根据所述物品类型与所述初始信息,确定目标物品的准确信息。
  10. 一种电子设备,包括存储器和处理器,所述存储器中存储有可在所述处理器上运行的计算机程序,其特征在于,所述处理器执行所述计算机程序时实现上述权利要求1至8任一项所述的方法的步骤。
PCT/CN2018/115497 2017-11-14 2018-11-14 物品识别方法、系统以及电子设备 WO2019096180A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711126214.3A CN107832795B (zh) 2017-11-14 2017-11-14 物品识别方法、系统以及电子设备
CN201711126214.3 2017-11-14

Publications (2)

Publication Number Publication Date
WO2019096180A1 true WO2019096180A1 (zh) 2019-05-23
WO2019096180A9 WO2019096180A9 (zh) 2019-08-29

Family

ID=61655311

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/115497 WO2019096180A1 (zh) 2017-11-14 2018-11-14 物品识别方法、系统以及电子设备

Country Status (2)

Country Link
CN (1) CN107832795B (zh)
WO (1) WO2019096180A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112685626A (zh) * 2019-10-17 2021-04-20 北京星选科技有限公司 一种对象类型的确定方法、装置、电子设备及存储介质

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107832795B (zh) * 2017-11-14 2021-07-27 深圳码隆科技有限公司 物品识别方法、系统以及电子设备
CN109190697A (zh) * 2018-08-29 2019-01-11 德比软件(上海)有限公司 一种酒店图片分类判断系统和方法
CN109409381B (zh) * 2018-09-18 2021-06-15 躺平设计家(北京)科技有限公司 基于人工智能的家具顶视图的分类方法及系统
CN112426100B (zh) * 2019-08-26 2022-10-21 美智纵横科技有限责任公司 一种控制方法、装置及存储介质
CN110737941A (zh) * 2019-10-12 2020-01-31 南京我爱我家信息科技有限公司 一种基于概率模型及像素统计模型的房屋装修程度识别系统和方法
CN116009560A (zh) * 2023-03-24 2023-04-25 科大讯飞股份有限公司 机器人及其寻桩方法、装置及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150178597A1 (en) * 2013-12-24 2015-06-25 Canon Kabushiki Kaisha Method of classifying objects in scenes
CN105426925A (zh) * 2015-12-28 2016-03-23 联想(北京)有限公司 图像标注方法和电子设备
CN106446933A (zh) * 2016-08-31 2017-02-22 河南广播电视大学 基于上下文信息的多目标检测方法
CN107194318A (zh) * 2017-04-24 2017-09-22 北京航空航天大学 目标检测辅助的场景识别方法
CN107832795A (zh) * 2017-11-14 2018-03-23 深圳码隆科技有限公司 物品识别方法、系统以及电子设备

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4253567B2 (ja) * 2003-03-28 2009-04-15 オリンパス株式会社 データオーサリング処理装置
CN102930246B (zh) * 2012-10-16 2015-04-08 同济大学 一种基于点云片段分割的室内场景识别方法
CN102945567B (zh) * 2012-10-19 2016-06-15 深圳先进技术研究院 室内场景的分类与重建方法及系统
CN104077784B (zh) * 2013-03-29 2018-02-27 联想(北京)有限公司 提取目标对象的方法和电子设备
CN115690558A (zh) * 2014-09-16 2023-02-03 华为技术有限公司 数据处理的方法和设备
CN104573711B (zh) * 2014-12-22 2017-12-15 上海交通大学 基于文本‑物体‑场景关系的物体和场景的图像理解方法
EP3115927A1 (en) * 2015-07-09 2017-01-11 Thomson Licensing Method and apparatus for processing a scene
CN106776619B (zh) * 2015-11-20 2020-09-04 百度在线网络技术(北京)有限公司 用于确定目标对象的属性信息的方法和装置
CN106650798B (zh) * 2016-12-08 2019-06-21 南京邮电大学 一种结合深度学习与稀疏表示的室内场景识别方法
CN106851098A (zh) * 2017-01-20 2017-06-13 努比亚技术有限公司 一种图像处理方法及移动终端
CN107341508B (zh) * 2017-06-22 2020-12-04 苏州飞搜科技有限公司 一种快速美食图片识别方法及系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150178597A1 (en) * 2013-12-24 2015-06-25 Canon Kabushiki Kaisha Method of classifying objects in scenes
CN105426925A (zh) * 2015-12-28 2016-03-23 联想(北京)有限公司 图像标注方法和电子设备
CN106446933A (zh) * 2016-08-31 2017-02-22 河南广播电视大学 基于上下文信息的多目标检测方法
CN107194318A (zh) * 2017-04-24 2017-09-22 北京航空航天大学 目标检测辅助的场景识别方法
CN107832795A (zh) * 2017-11-14 2018-03-23 深圳码隆科技有限公司 物品识别方法、系统以及电子设备

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112685626A (zh) * 2019-10-17 2021-04-20 北京星选科技有限公司 一种对象类型的确定方法、装置、电子设备及存储介质

Also Published As

Publication number Publication date
CN107832795B (zh) 2021-07-27
CN107832795A (zh) 2018-03-23
WO2019096180A9 (zh) 2019-08-29

Similar Documents

Publication Publication Date Title
WO2019096180A1 (zh) 物品识别方法、系统以及电子设备
JP6938116B2 (ja) 在庫管理装置および在庫管理方法
US9565521B1 (en) Automatic semantic labeling based on activity recognition
WO2020024691A1 (zh) 无人结算场景中异常检测方法、装置及设备
JP2019504379A (ja) 煙検出装置、方法及び画像処理装置
JP2016532917A (ja) ベニューマップの生成および更新
CN103049520A (zh) 利用照片来发起及施行动作
CN105187786A (zh) 智能猫眼的语音提醒方法及智能猫眼
US10895863B2 (en) Electronic device and method for controlling the same
US9615429B2 (en) Illuminating devices and systems
CN108764100B (zh) 一种目标行为检测方法及服务器
WO2013000404A1 (zh) 一种图像检测方法及装置
Saha et al. EnergyLens: Combining smartphones with electricity meter for accurate activity detection and user annotation
US10655951B1 (en) Determining relative positions of user devices
JP7113281B2 (ja) 冷蔵庫
KR102003691B1 (ko) 소품 등록 시스템
CN105405249A (zh) 防碰撞方法、装置、测距传感器及终端
JP2023093716A (ja) 冷蔵庫、管理システム、及びプログラム
CN114120221A (zh) 基于深度学习的环境核查方法、电子设备及存储介质
US20240095674A1 (en) Storage-retrieval management device, storage-retrieval management system, and storage-retrieval management method
US11657611B2 (en) Methods and systems for augmented reality room identification based on room-object profile data
US10215858B1 (en) Detection of rigid shaped objects
WO2019101002A1 (zh) 一种鞋的数据获取方法和设备以及计算机存储介质
US10540390B1 (en) Image-based item identification
US20240233384A9 (en) Storage-retrieval management device, storage-retrieval management system, and storage-retrieval management method

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18877894

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18877894

Country of ref document: EP

Kind code of ref document: A1