WO2019210575A1 - 不动产信息查询方法、装置、计算机设备和存储介质 - Google Patents

不动产信息查询方法、装置、计算机设备和存储介质 Download PDF

Info

Publication number
WO2019210575A1
WO2019210575A1 PCT/CN2018/095387 CN2018095387W WO2019210575A1 WO 2019210575 A1 WO2019210575 A1 WO 2019210575A1 CN 2018095387 W CN2018095387 W CN 2018095387W WO 2019210575 A1 WO2019210575 A1 WO 2019210575A1
Authority
WO
WIPO (PCT)
Prior art keywords
real estate
information
face photo
real
estate information
Prior art date
Application number
PCT/CN2018/095387
Other languages
English (en)
French (fr)
Inventor
李俊
冯心
罗均
莫智勋
熊恢
Original Assignee
平安科技（深圳）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 平安科技（深圳）有限公司 filed Critical 平安科技（深圳）有限公司
Publication of WO2019210575A1 publication Critical patent/WO2019210575A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
    • G06F21/31—User authentication
    • G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/16—Real estate
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/161—Detection; Localisation; Normalisation
    • G06V40/166—Detection; Localisation; Normalisation using acquisition arrangements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172—Classification, e.g. identification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/40—Spoof detection, e.g. liveness detection
    • G06V40/45—Detection of the body part being alive
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2221/00—Indexing scheme relating to security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F2221/21—Indexing scheme relating to G06F21/00 and subgroups addressing additional information or applications relating to security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F2221/2141—Access rights, e.g. capability lists, access control lists, access tables, access matrices

Definitions

  • the present application relates to the field of computer technology, and in particular, to a method, device, computer device and storage medium for inquiring real estate information.
  • the existing real estate information inquiry mainly includes two methods.
  • the first inquiry method is: the owner offline to the real estate registration center to make inquiries, which is time-consuming and laborious;
  • the second inquiry method is: the owner online inquiry, but in order to protect the queryer's privacy Only limited information can be queried when querying online. If you need to query the detailed real estate information, you need to return to the first query method above. Therefore, how to maximize the acquisition of real estate information while protecting the privacy of the inquirer is an urgent problem to be solved.
  • the main purpose of the present application is to provide a method, device, computer device and storage medium for inquiring real estate information for effectively managing privacy rights.
  • the present application provides a method for inquiring real estate information for inquiring about real estate information online at a place other than the real estate registration center, the method comprising: obtaining identity information of the inquirer;
  • the real estate information corresponding to the query authority is output.
  • the application also provides a real estate information inquiry device for performing online information inquiry on a place other than the real estate registration center, the device comprising:
  • An obtaining unit configured to obtain identity information of the querier
  • a photographing calling unit configured to take a first face photo of the inquirer, and call a first filing picture corresponding to the identity information to the public security system according to the identity information;
  • a calculating unit configured to calculate a similarity value between the first face photo and the first recorded picture
  • a searching unit configured to search, in the preset permission list, a query permission corresponding to the similarity value
  • an output unit configured to output real estate information corresponding to the query authority.
  • the application further provides a computer device comprising a memory and a processor, the memory storing computer readable instructions, the processor executing the computer readable instructions to implement the steps of any of the methods described above.
  • the present application also provides a computer non-transitory readable storage medium having stored thereon computer readable instructions, wherein the computer readable instructions are executed by a processor to implement the method of any of the above step.
  • the method, device, computer equipment and storage medium for querying real estate information of the present application are used for online real estate information inquiry in a place other than the real estate registration center, and the inquirer can query the real estate information without going to the real estate registration center, so that the inquirer can easily For example, banks, real estate agents, etc. handle the handling of real estate information business.
  • the identity verification of the querier can be performed. Only the verification can pass through to find the complete property information, and the finder can be maximized under the condition of protecting the privacy; if the finder uses the false information to register the real property information, At this time, the query cannot be successful, thereby reducing the occurrence of false registration.
  • FIG. 1 is a schematic flow chart of a method for querying real estate information according to an embodiment of the present application
  • FIG. 2 is a schematic flowchart of a method for querying real estate information according to an embodiment of the present application
  • FIG. 3 is a schematic block diagram showing the structure of a real estate information inquiry apparatus according to an embodiment of the present application.
  • FIG. 4 is a schematic block diagram showing the structure of a real estate information inquiry apparatus according to an embodiment of the present application.
  • FIG. 5 is a schematic block diagram showing the structure of a real estate information inquiry apparatus according to an embodiment of the present application.
  • FIG. 6 is a schematic block diagram showing the structure of a real estate information inquiry apparatus according to an embodiment of the present application.
  • FIG. 7 is a schematic block diagram showing the structure of a computing unit according to an embodiment of the present application.
  • FIG. 8 is a schematic block diagram showing the structure of a real estate information inquiry apparatus according to an embodiment of the present application.
  • FIG. 9 is a schematic block diagram showing the structure of a computer device according to an embodiment of the present application.
  • the present application provides a method for querying real estate information for performing online information inquiry on a place other than a real estate registration center, and the method includes:
  • S2 capturing a first face photo of the queryer, and calling a first file corresponding to the identity information to the public security system according to the identity information;
  • the above-mentioned real estate registration center is a place for supporting the work of real estate registration related policies, business, technology, and the legal affairs of land and resources, and it will set up corresponding offices in different administrative areas, such as Shenzhen real estate.
  • the registration center mainly handles the real estate registration business in Shenzhen.
  • the online information on the real estate information of the place other than the above-mentioned real estate registration center mainly refers to the office of the bank, real estate agency and other enterprises authorized by the real estate registration center, and the above-mentioned real estate information inquiry system is installed on the electronic equipment such as computers used.
  • the real estate information inquiry system is connected to the server of the real estate registration center through the interface of the http protocol. In this embodiment, only the electronic device such as a computer of an authorized enterprise can log in and use the above-mentioned real estate information inquiry system.
  • the real estate registration service system is generally installed on the office equipment.
  • the content of the real estate inquiry system presented on the screen of the computer is an online inquiry page, which is provided with an input window for inputting the identity information of the inquirer.
  • the identity information includes at least the ID number of the inquirer, and the basic information of the inquired property includes at least the specific address of the real estate.
  • the above identity information is generally manually entered into the real estate inquiry system by a bank or real estate agent.
  • the identity information of the inquirer may be reversely derived by inputting the basic information of the real estate to be inquired, that is, according to the basic information of the real estate to be inspected, the real estate center is searched for the corresponding property owner information, and the property owner information is Is the default querier information.
  • the real estate information inquiry system controls the camera attached thereto to take a photo of the face of the inquirer (generally including a positive face), that is, the first face photo described above.
  • the first record picture mentioned above refers to the picture when the inquirer handles the identity card.
  • the real estate information inquiry system is a system authorized by the public security system, and the public security system allows the real estate information inquiry system to access the database in which the identity card is stored, so the real estate information inquiry system can be based on the obtained identification number to the public security system.
  • the ID picture corresponding to the identity information is called, and the face picture on the ID picture is the first recorded picture.
  • the calculation process of the similarity value is specifically: inputting the first face photo into the preset face recognition model to perform the operation to output the feature vector of the first face photo, and the first record picture Inputting into the above-mentioned face recognition model, the feature vector of the first record picture is outputted, and then the feature vector of the first face photo and the feature vector of the first record picture are similarly calculated.
  • Specific similarity values can be calculated using Euclidian Distance, Manhattan Distance (Manhattan) Distance), Minkowski distance or cosine similarity is calculated by an algorithm.
  • the preset permission list is a one-to-one mapping list, wherein there is a column of similarity values and a column of query permissions, and each similarity value corresponds to one query permission.
  • the similarity value calculated by the similarity value in the above step S3 cannot be found in the permission list
  • the similarity between the similarity value calculated in the above step S3 and the smallest in the above-mentioned step S3 is selected.
  • the value is used. For example, the calculated similarity value of the first face photo and the first recorded image is 10, and the permission list has no similarity value of 10, and only the similarity values of 3, 6, 9, and 12 are obtained.
  • the query authority selects a similarity value 9 close to the similarity value of 10, and then select the query authority corresponding to the similarity value of 9.
  • the above query authority refers to the query authority of the queryer. The higher the query authority, the more real estate information that can be found. If the query permission is 100%, you can view all the information of the real estate to be checked.
  • the information of the real estate to be checked is retrieved according to the query authority to the real estate registration center.
  • the output method includes a variety of methods, including printing directly through a printer connected to a computer equipped with a real estate information inquiry system; or, displaying through a computer screen; or, by playing in a voice broadcast form, the manner of voice broadcast is beneficial to the inquiry.
  • the step of obtaining the identity information of the inquirer may be obtained by means of voice recognition, such as receiving the voice input of the inquirer, and then converting the voice input into text.
  • the document with the official seal of the real estate center may be generated for the content to be used for use.
  • the inquirer needs to perform mortgage on the property when the bank handles the loan and the like, and the bank staff can accurately view the real estate information of the inquirer on the spot.
  • the bank staff logs in the real estate information inquiry system, and then logs in using the staff member's account, and after inputting the identity information of the inquirer, collects the current face information of the inquirer, and combines the record of the inquirer in the public security system. Determine whether the queryer is the owner of the property to be inquired, and if so, open all the query permissions to display all the information of the real estate to be inspected.
  • the intermediary worker when the inquirer purchases a house or the like in the real estate agent, it is necessary to check whether the inquirer meets the local purchase conditions, and the intermediary worker can accurately view the inquirer's real estate information on the spot to determine the Whether the inquirer can purchase a house, specifically, the intermediary worker logs in to the real estate information inquiry system, and then logs in using the staff member's account, and after inputting the identity information of the inquirer, collects the current face information of the inquirer, and combines the inquirer with The record of the public security system determines whether the inquirer is the identity of the registered identity information, and if so, displays all the real estate information of the inquirer. Further, it is also possible to link to the CPF account or the social security account, to view the information of the inquirer's provident fund and social security payment, and further facilitate the judgment of the intermediary staff.
  • the method before the step S5 of outputting the real estate information corresponding to the query authority, the method includes:
  • the real information is the information that the real estate to be checked is registered in the real estate registration center, and the information of the real estate to be inspected at this time is all that is not registered in the real estate registration center corresponding to the real estate to be inspected.
  • Information about privacy information is the content extracted from the information registered by the above-mentioned real estate registration center, and may be randomly selected or extracted according to a specified rule, such as extracting information about the transfer in the real estate information.
  • the real information is obtained, the meaning of the real information is obtained according to the voice analysis method, and then the approximate error information is generated according to the meaning. For example, in the information of the real estate to be checked, the real transfer record is transferred on October 15, 2010.
  • step S5 After extracting the part, it generates error information such as the transfer on October 15, 2011, the transfer on November 15, 2010, and the transfer on October 25, 2010.
  • error information such as the transfer on October 15, 2011, the transfer on November 15, 2010, and the transfer on October 25, 2010.
  • the above correct information and three error messages are formed.
  • the four alternative answers to the multiple-choice question "Select the answer to the correct transfer record of the real estate" are selected by the queryer, and the action of the above step S5 is performed only if the queryer answers correctly.
  • the step of answering the multiple-choice question can be further determined whether the inquirer is the property owner of the real property to be inspected, so as to effectively prevent the private information about the real property to be inspected from being exposed.
  • multiple similar multiple-choice questions can be generated, and the action of step S5 described above is performed only after the queryer continuously answers the specified number of multiple-choice questions.
  • the method before the step S5 of outputting the real estate information corresponding to the query authority, the method includes:
  • S501 Determine, by the first face photo, the first age of the querier, and obtain the second age of the querier by using the identity information.
  • the first age is the age estimated by the real estate inquiry system according to the first face photo, and the specific method is that the first face photo is input into the preset age recognition model, and the operation is performed. Output the above first age.
  • the above age recognition model is based on a model trained by a convolutional neural network. First, a large number of face photos are acquired, and the corresponding real age is marked on each face photo, and then input into a preset convolutional neural network for training. After the training is finished, another batch of face photos with the true age of the mark is verified, and if the verification is passed, the above-mentioned age recognition model is obtained.
  • the second age mentioned above is an age calculated based on the identity information, such as obtaining the date of birth based on the entered ID number, and then calculating the second age in combination with the current date. If the current inquirer is a real property owner, then the first age and the second age should be the same. Considering the accuracy of the model based judgment, set an error range (the above preset range), only the first age and the second. If the difference between the ages is within the error range, it will be considered that the inquirer may be the property owner of the real estate to be inspected, and then the action of the above step S5 is performed to effectively prevent the privacy information about the real estate to be inspected from being exposed.
  • an error range the above preset range
  • a "command that does not output the query result" may be generated, the real estate information query system is controlled to stop outputting the query result, or a smaller query right is generated. To output less information about the real estate to be inspected.
  • the method before the step S3 of calculating the similarity value between the first face photo and the first recorded picture, the method includes:
  • the calculation result of the similarity between the first face photo and the first recorded picture is affected, so the first face photo is compared before the comparison.
  • the beauty effect is removed from the first face photo.
  • the specific process is: first input the first face photo into the racial recognition model, determine the race of the inquirer, that is, judge the inquirer to be a yellow race, a black race or a caucasian, etc., and then determine the query according to the race. The true eye color of the person, if the judge is Yellow Bell, then the eyes should be black. If the color of the human eye in the first face photo is other colors, modify it to black.
  • the above-mentioned racial recognition model is a model obtained by training with a convolutional neural network.
  • the training method is a conventional training using the training set, and the test set is tested, etc., and will not be described here.
  • the pre-processed first face photo is used to obtain an accurate similarity value between the first face photo and the first recorded picture.
  • the first face photo may include a left view first face photo, a right view first face photo, and a front view first face photo; and the calculating the first face photo Step S3 of the similarity value with the first record picture includes:
  • the left-view first face photo and the right-view first face photo are photographs taken from the left and right sides of the face of the inquirer, respectively. Since the left-view first face photo and the right-view first face photo cannot cover all of the facial features, the similarity value with the first recorded image is lower relative to the frontal first face photo.
  • the left-view first face photo, the right-view first face photo, and the front-view first face photo are respectively compared with the first recorded image, thereby obtaining three sets of sub-similarity values, but not The three sets of sub-similarity values are directly averaged, but are weighted average calculations.
  • the sub-similarity value corresponding to the first-view photo of the left-view is multiplied by a percentage
  • the sub-similarity value corresponding to the first-view photo of the right-view is multiplied. Multiply a percentage by the percentage similarity and the sub-similarity value corresponding to the first face photo, and then add to obtain the final similarity value.
  • the face image of the inquirer at different angles is taken to prevent the inquirer of the non-proprietor from using the photo of the owner (the photo can only display a fixed direction image, such as a photo on the front of the face, the front photo cannot be used. Take a picture of the face on the left and right sides) Deceive the real estate information inquiry system and obtain all the information of the real estate to be inspected.
  • the method includes:
  • S7 Predict the real estate buying and selling behavior of the queryer according to the frequency and the geographical distribution.
  • the real estate information inquiry system uploads the recorded data to a server, and the recorded content includes the query time of the queryer, where the query was made, and the like. These records are predicted according to a preset method, and the authenticity of the inquirer can be judged.
  • a queryer queries real estate agents in different regions of Shenzhen within three months, indicating that they have contacted different intermediaries in different regions, and the intermediary introduces the listings to them. In the past, the purchaser's purchase conditions are generally verified. At this time, the real estate information inquiry system is required to query the real estate information of the inquirer, thereby determining whether the inquirer has the purchase qualification.
  • the inquiring person can send relevant information to help the inquiring person to timely understand the real estate information he cares about.
  • the method includes:
  • the subject in the shooting area is subjected to living body detection to prevent others from using the photo fraud system of the property owner corresponding to the real estate information.
  • the manner of performing living body detection on the subject in the shooting area includes various methods, such as determining the "motion" of the respective pixel positions by using the temporal variation and correlation of the pixel intensity data in the image sequence according to the optical flow method, from the image sequence.
  • the operation information of each pixel is obtained, and Gaussian difference filter, LBP (Local Binary Pattern, which is an operator for describing local features of the image) and support vector machine are used for statistical analysis of data.
  • LBP Local Binary Pattern, which is an operator for describing local features of the image
  • This method of living detection can achieve blind testing without the help of the inquirer.
  • the 3D camera is used to capture the face, and the corresponding 3D data of the face area is obtained, and further analysis is performed based on the data, and finally whether the face is from a living body or a non-living body.
  • the sources of non-living are relatively wide, including photos and videos of media such as mobile phones and Pads, and various photos of different materials (including bending, folding, cropping, digging, etc.).
  • the most discriminative feature is selected to train the classifier, and the trained classifier is used to distinguish between live and non-living.
  • the method includes: collecting each open query of each login account of the real estate information inquiry system. Permission; within a specified length of time, if the ratio of the number of times the query permission is not fully open to the total number of queries is greater than the preset ratio, a reminder message is generated.
  • the reminder information includes a form such as mail, and the mail is generally sent to the designated leader mailbox.
  • the bank or intermediary has a corresponding account, and the number of times the inquirer and the recorded photo are inconsistent in the face recognition process (there is no full open access permission), the number of inconsistencies will be a than the total number of times identified b, if the ratio is greater than the preset ratio, it means that the staff of the account is not serious in the work, and can not clearly understand the true information of the inquirer in the process of handling the business, that is, there is no If the communication is clear, the action of real estate information inquiry will be carried out, and the business level needs to be improved.
  • the online The pursuit personnel library searches for the identity information that matches the querier; if the identity information matching the querier is found in the online escaping personnel library, the automatic alarm is automatically generated. Since the first face photo is the same person as the first recorded picture, it can be determined that the inquirer is a valid and valid citizen. At this time, the real estate information inquiry system will automatically link to the online pursuit personnel database to find the escaped personnel information that matches the ID card information of the inquirer. If the same escaped personnel information is found, the inquirer is an online pursuit personnel. The system will send an alarm signal in the background.
  • the alarm signal may be an administrator who notifies the real estate information inquiry system.
  • the real estate information inquiry system is installed at a bank's venue, and the alarm signal may be sent to the bank security department through the background.
  • the 110 alarm platform can also be directly connected through the background, and the related information such as the address information is sent to the 110 alarm platform and the like.
  • the real estate information inquiry method of the embodiment is used for inquiring the real estate information online in a place other than the real estate registration center, and the inquirer can query the real estate information without going to the real estate registration center, so that the inquirer can go to a bank, a real estate agent, etc. Handling when real estate information business is required.
  • the identity verification of the querier can be performed. Only the verification can pass through to find the complete property information, and the finder can be maximized under the condition of protecting the privacy; if the finder uses the false information to register the real property information, At this time, the query cannot be successful, thereby reducing the occurrence of false registration.
  • the present application provides a real estate information inquiry device for performing real-time information online inquiry at a place other than a real estate registration center, and the device includes:
  • the obtaining unit 10 is configured to obtain identity information of the querier
  • the photographing invoking unit 20 is configured to capture a first face photo of the inquirer, and call a first filing picture corresponding to the identity information to the public security system according to the identity information;
  • the calculating unit 30 is configured to calculate a similarity value between the first face photo and the first recorded picture
  • the searching unit 40 is configured to search, in the preset permission list, a query permission corresponding to the similarity value
  • the output unit 50 is configured to output real estate information corresponding to the query authority.
  • the above-mentioned real estate registration center is a place for supporting the work of real estate registration related policies, business, technology, and the legal affairs of land and resources, and it will set up corresponding offices in different administrative areas, such as Shenzhen real estate.
  • the registration center mainly handles the real estate registration business in Shenzhen.
  • the online information on the real estate information of the place other than the above-mentioned real estate registration center mainly refers to the office of the bank, real estate agency and other enterprises authorized by the real estate registration center, and the above-mentioned real estate information inquiry system is installed on the electronic equipment such as computers used.
  • the real estate information inquiry system is connected to the server of the real estate registration center through the interface of the http protocol. In this embodiment, only the electronic device such as a computer of an authorized enterprise can log in and use the above-mentioned real estate information inquiry system.
  • the real estate registration service system is generally installed on the office equipment.
  • the content of the real estate inquiry system presented on the screen of the computer is an online inquiry page, which is provided with an input window for inputting the identity information of the inquirer.
  • the identity information includes at least the ID number of the inquirer.
  • the above identity information is generally manually entered into the real estate inquiry system by a bank or real estate agent.
  • the identity information of the inquirer may be reversely derived by inputting the basic information of the real estate to be inquired, that is, according to the basic information of the real estate to be inspected, the real estate center is searched for the corresponding property owner information, and the property owner information is Is the default querier information.
  • the above-mentioned real estate information inquiry system controls the photograph of the face of the inquirer (usually including the face), that is, the first face photograph described above.
  • the first record picture mentioned above refers to the picture when the inquirer handles the identity card.
  • the real estate information inquiry system is a system authorized by the public security system, and the public security system allows the real estate information inquiry system to access the database in which the identity card is stored, so the real estate information inquiry system can be based on the obtained identification number to the public security system.
  • the ID picture corresponding to the identity information is called, and the face picture on the ID picture is the first recorded picture.
  • the calculating process of the similarity value is specifically: inputting the first face photo into the preset face recognition model to perform the operation to output the feature vector of the first face photo, and the first recorded image Inputting into the above-mentioned face recognition model, the feature vector of the first record picture is outputted, and then the feature vector of the first face photo and the feature vector of the first record picture are similarly calculated.
  • Specific similarity values can be calculated using Euclidian Distance, Manhattan Distance (Manhattan) Distance), Minkowski distance or cosine similarity is calculated by an algorithm.
  • the preset permission list is a one-to-one mapping list, wherein there is a column of similarity values and a column of query permissions, and each similarity value corresponds to one query authority.
  • the similarity value calculated by the calculating unit 30 is not found in the permission list, the similarity value of the difference between the similarity values calculated in the permission list and the above-mentioned Liu table is selected. use. For example, the calculated similarity value of the first face photo and the first recorded image is 10, and the permission list has no similarity value of 10, and only the similarity values of 3, 6, 9, and 12 are obtained.
  • select a similarity value 9 close to the similarity value of 10 and then select the query authority corresponding to the similarity value of 9.
  • the above query authority refers to the query authority of the queryer. The higher the query authority, the more real estate information that can be found. If the query permission is 100%, you can view all the information of the real estate to be checked.
  • the output unit 50 it is a unit that retrieves the information of the real estate to be inspected according to the inquiry authority to the real estate registration center.
  • the output method includes a variety of methods, including printing directly through a printer connected to a computer equipped with a real estate information inquiry system; or, displaying through a computer screen; or, by playing in a voice broadcast form, the manner of voice broadcast is beneficial to the inquiry.
  • the method for obtaining the identity information of the inquirer may be obtained by means of voice recognition, such as receiving the voice input of the inquirer, and then converting the voice input into text.
  • the document with the official seal of the real estate center may be generated for the content to be used for use.
  • the inquirer needs to perform mortgage on the property when the bank handles the loan and the like, and the bank staff can accurately view the real estate information of the inquirer on the spot.
  • the bank staff logs in the real estate information inquiry system, and then logs in using the staff member's account, and after inputting the identity information of the inquirer, collects the current face information of the inquirer, and combines the record of the inquirer in the public security system. Determine whether the queryer is the owner of the property to be inquired, and if so, open all the query permissions to display all the information of the real estate to be inspected.
  • the intermediary worker when the inquirer purchases a house or the like in the real estate agent, it is necessary to check whether the inquirer meets the local purchase conditions, and the intermediary worker can accurately view the inquirer's real estate information on the spot to determine the Whether the inquirer can purchase a house, specifically, the intermediary worker logs in to the real estate information inquiry system, and then logs in using the staff member's account, and after inputting the identity information of the inquirer, collects the current face information of the inquirer, and combines the inquirer with The record of the public security system determines whether the inquirer is the identity of the registered identity information, and if so, displays all the real estate information of the inquirer. Further, it is also possible to link to the CPF account or the social security account, to view the information of the inquirer's provident fund and social security payment, and further facilitate the judgment of the intermediary staff.
  • the real estate information query device includes:
  • the extracting unit 51 is configured to extract part of the real estate information from the real estate information as real information
  • a generating unit 52 configured to generate error information that is similar to the real information
  • the mixing unit 53 is configured to generate the multiple choice question by using the real information and the error information for the querier to select;
  • the receiving generating unit 54 is configured to execute the step of outputting the real estate information corresponding to the query right if the answer selected by the queryer is received.
  • the real information of the real estate is the information that the real estate to be checked is registered in the real estate registration center, and the information of the real estate to be inspected at this time is corresponding. All information that does not involve private information registered with the real estate registration center of the real estate to be inspected.
  • the real information is the content extracted from the information registered by the above-mentioned real estate registration center, and may be randomly selected or extracted according to a specified rule, such as extracting information about the transfer in the real estate information.
  • a specified rule such as extracting information about the transfer in the real estate information.
  • the real transfer record is transferred on October 15, 2010. After extracting the part, it generates error information such as the transfer on October 15, 2011, the transfer on November 15, 2010, and the transfer on October 25, 2010.
  • the above correct information and three error messages are formed.
  • the four alternative answers to the multiple-choice question "Choose the answer to the correct transfer record of the real estate" are for the finder to select, and only the finder's answer is correct, and the real estate information corresponding to the query authority will be output.
  • the step of answering the multiple-choice question can be further determined whether the inquirer is the property owner of the real property to be inspected, so as to effectively prevent the private information about the real property to be inspected from being exposed.
  • multiple similar multiple-choice questions can be generated, and the real estate information corresponding to the query authority is output only after the queryer continuously answers the specified number of multiple-choice questions.
  • the foregoing real estate information querying apparatus includes:
  • the age obtaining unit 501 is configured to determine, by using the first face photo, the first age of the querier, and obtain the second age of the querier by using the identity information;
  • the determining unit 502 is configured to determine whether the difference between the first age and the second age is within a preset range
  • the generating instruction unit 503 is configured to: if the difference between the first age and the second age is within a preset range, perform a step of outputting real estate information corresponding to the query authority.
  • the first age is the age estimated by the real estate inquiry system according to the first face photo
  • the specific method is: inputting the first face photo into the preset In the age recognition model, the first age is output through the operation.
  • the above age recognition model is based on a model trained by a convolutional neural network. First, a large number of face photos are acquired, and the corresponding real age is marked on each face photo, and then input into a preset convolutional neural network for training. After the training is finished, another batch of face photos with the true age of the mark is verified, and if the verification is passed, the above-mentioned age recognition model is obtained.
  • the second age mentioned above is an age calculated based on the identity information, such as obtaining the date of birth based on the entered ID number, and then calculating the second age in combination with the current date. If the current inquirer is a real property owner, then the first age and the second age should be the same. Considering the accuracy of the model based judgment, set an error range (the above preset range), only the first age and the second. If the difference between the ages is within the error range, it will be considered that the inquirer may be the property owner of the real estate to be inspected, and then the real estate information corresponding to the inquiry authority is outputted to effectively prevent the privacy information about the real estate to be inspected from being exposed.
  • an error range the above preset range
  • a "command that does not output the query result" may be generated, the real estate information query system is controlled to stop outputting the query result, or a smaller query right is generated. To output less information about the real estate to be inspected.
  • the foregoing real estate information querying apparatus includes:
  • a determining unit 31 configured to input the first face photo into a preset racial recognition model, and determine a racial attribution of the inquirer to determine a true eye color of the inquirer;
  • the color determining unit 32 is configured to determine whether the color of the human eye in the first face photo is the same as the true eye color
  • the reducing unit 33 is configured to restore the color of the human eye in the first face photo to the real eye color if the color of the human eye in the first face photo is different from the real eye color, so as to facilitate A similarity comparison is performed with the first recorded picture.
  • the beauty effect is removed from the first face photo.
  • the specific process is: first input the first face photo into the racial recognition model, determine the race of the inquirer, that is, judge the inquirer to be a yellow race, a black race or a caucasian, etc., and then determine the query according to the race. The true eye color of the person, if the judge is Yellow Bell, then the eyes should be black. If the color of the human eye in the first face photo is other colors, modify it to black.
  • the above-mentioned racial recognition model is a model obtained by training with a convolutional neural network.
  • the training method is a conventional training using the training set, and the test set is tested, etc., and will not be described here.
  • the pre-processed first face photo is used to obtain an accurate similarity value between the first face photo and the first recorded picture.
  • the first face photo may include a left-view first face photo, a right-view first face photo, and a front-view first face photo;
  • the calculating unit 30 includes :
  • the comparison module 301 is configured to compare the left-view first face photo, the right-view first face photo, and the front-view first face photo with the first recorded image, respectively. Obtaining three sets of sub-similarity values corresponding to the left-view first face photo, the right-view first face photo, and the front-view first face photo respectively;
  • the weighting calculation module 302 is configured to calculate the similarity values of the first face photo and the first record picture by calculating the obtained three sets of sub-similarity values according to a preset weighted average rule.
  • the left-view first face photo and the right-view first face photo are photos taken from the left and right sides of the face of the inquirer, respectively. Since the left-view first face photo and the right-view first face photo cannot cover all of the facial features, the similarity value with the first recorded image is lower relative to the frontal first face photo.
  • the left-view first face photo, the right-view first face photo, and the front-view first face photo are respectively compared with the first recorded image, thereby obtaining three sets of sub-similarity values, but not The three sets of sub-similarity values are directly averaged, but are weighted average calculations.
  • the sub-similarity value corresponding to the first-view photo of the left-view is multiplied by a percentage
  • the sub-similarity value corresponding to the first-view photo of the right-view is multiplied. Multiply a percentage by the percentage similarity and the sub-similarity value corresponding to the first face photo, and then add to obtain the final similarity value.
  • the face image of the inquirer at different angles is taken to prevent the inquirer of the non-proprietor from using the photo of the owner (the photo can only display a fixed direction image, such as a photo on the front of the face, the front photo cannot be used. Take a picture of the face on the left and right sides) Deceive the real estate information inquiry system and obtain all the information of the real estate to be inspected.
  • the real estate information querying device further includes:
  • a collecting unit 60 configured to collect the frequency and geographical distribution of the finder to query its real estate
  • the predicting unit 70 is configured to predict, according to the frequency and the geographical distribution, the real estate buying and selling behavior of the queryer.
  • the real estate information inquiry system uploads the recorded data to a server, and the recorded content includes the query time of the queryer, where the query has been made, and the like. These records are predicted according to a preset method, and the authenticity of the inquirer can be judged.
  • a queryer queries real estate agents in different regions of Shenzhen within three months, indicating that they have contacted different intermediaries in different regions, and the intermediary introduces the listings to them. In the past, the purchaser's purchase conditions are generally verified. At this time, the real estate information inquiry system is required to query the real estate information of the inquirer, thereby determining whether the inquirer has the purchase qualification.
  • the inquiring person can send relevant information to help the inquiring person to timely understand the real estate information he cares about.
  • the real estate information querying device further includes:
  • a living body detecting unit for performing living body detection on a subject in a shooting area
  • the living body determining unit is configured to determine that the first face photo is a usable photo if the detection result is that the subject is a living body.
  • the manner of performing living body detection on the subject in the shooting area includes various methods, such as determining the "motion" of the respective pixel positions by using the temporal variation and correlation of the pixel intensity data in the image sequence according to the optical flow method, from the image sequence.
  • the operation information of each pixel is obtained, and Gaussian difference filter, LBP (Local Binary Pattern, which is an operator for describing local features of the image) and support vector machine are used for statistical analysis of data.
  • LBP Local Binary Pattern, which is an operator for describing local features of the image
  • This method of living detection can achieve blind testing without the help of the inquirer.
  • the 3D camera is used to capture the face, and the corresponding 3D data of the face area is obtained, and further analysis is performed based on the data, and finally whether the face is from a living body or a non-living body.
  • the sources of non-living are relatively wide, including photos and videos of media such as mobile phones and Pads, and various photos of different materials (including bending, folding, cropping, digging, etc.).
  • the most discriminative feature is selected to train the classifier, and the trained classifier is used to distinguish between live and non-living.
  • the real estate information querying device further includes: a collection authority unit, configured to collect each open query permission of each login account of the real estate information inquiry system; and a reminding unit, configured to be within a specified length of time, If the ratio of the number of times the query permission is not completely open to the total number of queries is greater than the preset ratio, a reminder message is generated.
  • the reminder information includes a form such as mail, and the mail is generally sent to the designated leader mailbox.
  • the bank or intermediary has a corresponding account, and the number of times the inquirer and the recorded photo are inconsistent in the face recognition process (there is no full open access permission), the number of inconsistencies will be a than the total number of times identified b, if the ratio is greater than the preset ratio, it means that the staff of the account is not serious in the work, and can not clearly understand the true information of the inquirer in the process of handling the business, that is, there is no If the communication is clear, the action of real estate information inquiry will be carried out, and the business level needs to be improved.
  • the online The pursuit personnel library searches for the identity information that matches the querier; if the identity information matching the querier is found in the online escaping personnel library, the automatic alarm is automatically generated. Since the first face photo is the same person as the first recorded picture, it can be determined that the inquirer is a valid and valid citizen. At this time, the real estate information inquiry system will automatically link to the online pursuit personnel database to find the escaped personnel information that matches the ID card information of the inquirer. If the same escaped personnel information is found, the inquirer is an online pursuit personnel. The system will send an alarm signal in the background.
  • the alarm signal may be an administrator who notifies the real estate information inquiry system.
  • the real estate information inquiry system is installed at a bank's venue, and the alarm signal may be sent to the bank security department through the background.
  • the 110 alarm platform can also be directly connected through the background, and the related information such as the address information is sent to the 110 alarm platform and the like.
  • the real estate information inquiry device of the embodiment is used for inquiring the real estate information online at a place other than the real estate registration center, and the inquirer can query the real estate information without going to the real estate registration center, so that the inquirer can go to a bank, a real estate agent, etc. Handling when real estate information business is required.
  • the identity verification of the querier can be performed. Only the verification can pass through to find the complete property information, and the finder can be maximized under the condition of protecting the privacy; if the finder uses the false information to register the real property information, At this time, the query cannot be successful, thereby reducing the occurrence of false registration.
  • the computer device may be a server, and its internal structure may be as shown in FIG. 9.
  • the computer device includes a processor, memory, network interface, and database connected by a system bus. Among them, the computer designed processor is used to provide calculation and control capabilities.
  • the memory of the computer device includes a non-volatile storage medium, an internal memory.
  • the non-volatile storage medium stores an operating system, computer readable instructions, and a database.
  • the memory provides an environment for the operation of operating systems and computer readable instructions in a non-volatile storage medium.
  • the database of the computer device is used to store data such as a face recognition model.
  • the network interface of the computer device is used to communicate with an external terminal via a network connection.
  • the computer readable instructions are executed by a processor to implement the processes of the various method embodiments described above.
  • the embodiment of the present application further provides a computer non-volatile readable storage medium, on which computer readable instructions are stored, and when the computer readable instructions are executed by the processor, the flow of the foregoing method embodiments is implemented.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Multimedia (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Computer Interaction (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Business, Economics & Management (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Tourism & Hospitality (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Security & Cryptography (AREA)
  • Data Mining & Analysis (AREA)
  • Computer Hardware Design (AREA)
  • Software Systems (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Economics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Collating Specific Patterns (AREA)

Abstract

一种不动产信息查询方法、装置、计算机设备和存储介质，获取查询者的身份信息（S1）；拍摄查询者的第一人脸照片，以及根据身份信息到公安系统调用与身份信息对应的第一备案图片（S2）；计算第一人脸照片与第一备案图片的相似度值（S3）；在预设的权限列表中查找与相似度值对应的查询权限（S4）；输出与查询权限对应的不动产信息（S5）。

Description

不动产信息查询方法、装置、计算机设备和存储介质
本申请要求于2018年5月4日提交中国专利局、申请号为2018104220899，申请名称为“不动产信息查询方法、装置、计算机设备和存储介质”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
本申请涉及到计算机技术领域，特别是涉及到一种不动产信息查询方法、装置、计算机设备和存储介质。
 背景技术
现有的不动产信息查询主要包括两种方式，第一种查询方式是：业主线下到不动产登记中心进行查询，费时费力；第二种查询方式是：业主线上查询，但是为了保护查询者隐私，在线上查询时只能查询到有限的信息。如果需要查询详细的不动产信息，又要回归到上述的第一种查询方式。所以，如何在保护查询者隐私的情况下最大限度地获取不动产信息，是亟需解决的问题。
 技术问题
本申请的主要目的为提供一种有效管理隐私权限的不动产信息查询方法、装置、计算机设备和存储介质。
 技术解决方案
本申请提出一种不动产信息查询方法，用于在不动产登记中心之外的场所进行不动产信息线上查询，所述方法包括：获取查询者的身份信息；
拍摄所述查询者的第一人脸照片，以及根据所述身份信息到公安系统调用与所述身份信息对应的第一备案图片；
计算所述第一人脸照片与所述第一备案图片的相似度值；
在所述预设的权限列表中查找与所述相似度值对应的查询权限；
输出与所述查询权限对应的不动产信息。
本申请还提供一种不动产信息查询装置，用于在不动产登记中心之外的场所进行不动产信息线上查询，所述装置包括：
获取单元，用于获取查询者的身份信息；
拍摄调用单元，用于拍摄所述查询者的第一人脸照片，以及根据所述身份信息到公安系统调用与所述身份信息对应的第一备案图片；
计算单元，用于计算所述第一人脸照片与所述第一备案图片的相似度值；
查找单元，用于在所述预设的权限列表中查找与所述相似度值对应的查询权限；
输出单元，用于输出与所述查询权限对应的不动产信息。
本申请还提供一种计算机设备，包括存储器和处理器，所述存储器存储有计算机可读指令，所述处理器执行所述计算机可读指令时实现上述任一项所述方法的步骤。
本申请还提供一种计算机非易失性可读存储介质，其上存储有计算机可读指令，其特征在于，所述计算机可读指令被处理器执行时实现上述任一项所述的方法的步骤。
 有益效果
本申请的不动产信息查询方法、装置、计算机设备和存储介质，用于在不动产登记中心之外的场所进行不动产信息线上查询，查询者无须到不动产登记中心即可查询房产信息，方便查询者到如银行、房产中介等办理需要不动产信息业务时的办理。通过人脸识别技术，进行查询者的身份验证，只有验证通过才能查询到完整的房产信息，在保护隐私的条件下，最大程度地方便了查询者；如果查询者使用虚假信息进行不动产信息登记，此时则无法查询成功，进而降低虚假登记的情况发生。
 附图说明
图1 为本申请一实施例的不动产信息查询方法的流程示意图；
图2 为本申请一实施例的不动产信息查询方法的流程示意图；
图3 为本申请一实施例的不动产信息查询装置的结构示意框图；
图4 为本申请一实施例的不动产信息查询装置的结构示意框图；
图5 为本申请一实施例的不动产信息查询装置的结构示意框图；
图6 为本申请一实施例的不动产信息查询装置的结构示意框图；
图7 为本申请一实施例的计算单元的结构示意框图；
图8 为本申请一实施例的不动产信息查询装置的结构示意框图；
图9 为本申请一实施例的计算机设备的结构示意框图。
 本发明的最佳实施方式
参照图1，本申请提供一种不动产信息查询方法，用于在不动产登记中心之外的场所进行不动产信息线上查询，所述方法包括：
S1、获取查询者的身份信息；
S2、拍摄所述查询者的第一人脸照片，以及根据所述身份信息到公安系统调用与所述身份信息对应的第一备案图片；
S3、计算所述第一人脸照片与所述第一备案图片的相似度值；
S4、在所述预设的权限列表中查找与所述相似度值对应的查询权限；
S5、输出与所述查询权限对应的不动产信息。
本实施例中，上述不动产登记中心是承担不动产登记有关政策、业务、技术等的支撑工作以及国土资源法律事务工作的场所，其会在不同的行政区域设置对应的办事处等，比如深圳市不动产登记中心，其主要办理深圳市的不动产登记业务等。上述不动产登记中心之外的场所进行不动产信息线上查询，主要是指被不动产登记中心授权的银行、房产中介等企业的办公场所，其使用的电脑等电子设备上安装有上述不动产信息查询系统，不动产信息查询系统是通过http协议的接口与不动产登记中心的服务器连接。本实施例中，只有授权的企业的电脑等电子设备才可以登录使用上述不动产信息查询系统。
如上述步骤S1所述，一般在办公设备上安装不动产登记业务系统，上述不动产查询系统呈现在计算机的屏幕上的内容是一个线上查询页面，其设置有输入查询者身份信息的输入窗口等。其中，身份信息至少包括查询者的身份证号，待查询房产基本信息至少包括房产的具体地址。本申请中，上述身份信息一般是由银行或房产中介的工作人员手动输入到不动产查询系统中。在其它实施例中，还可以通过输入待查不动产的基本信息反向推导出查询者的身份信息，即根据待查不动产的基本信息到不动产中心查找与其对应的产权人信息，该产权人信息即为默认的查询者信息。
如上述步骤S2所述，上述不动产信息查询系统会控制与其连接的摄像头拍摄查询者的脸部照片（通常包括正脸照），即上述的第一人脸照片。上述第一备案图片是指查询者办理身份证时的图片。本实施例中，上述不动产信息查询系统是公安系统授权的系统，公安系统允许不动产信息查询系统访问其存储有身份证的数据库，因此不动产信息查询系统可以根据获取到的身份证号到公安系统中调用与所述身份信息对应的身份证图片，而身份证图片上有人脸图片，该人脸图片即为上述的第一备案图片。
如上述步骤S3所述，相似度值的计算过程具体为：将第一人脸照片输入到预设的人脸识别模型中进行运算输出第一人脸照片的特征向量，以及将第一备案图片输入到上述的人脸识别模型中进行运算输出第一备案图片的特征向量，然后将第一人脸照片的特征向量和第一备案图片的特征向量进行相似度计算。具体的相似度值计算可以使用欧几里得距离（Eucledian Distance）、曼哈顿距离（Manhattan Distance）、明可夫斯基距离（Minkowski distance）或者余弦相似度中一种算法进行计算。
如上述步骤S4所述，上述预设的权限列表是一个一对一的映射列表，其中有相似度值的一列，以及查询权限的一列，每一个相似度值对应一个查询权限。在本实施例中，当在权限列表无法查找到与上述步骤S3中计算出相似度值相同的相似度值时，则选择权限列表中与上述步骤S3中计算出相似度值差值最小的相似度值使用。比如，计算出的第一人脸照片与所述第一备案图片的相似度值为10，而权限列表中没有相似度值为10的选项，而只有3、6、9、12等相似度值选项，那么选择与相似度值10接近的相似度值9，然后选择相似度值9对应的查询权限。上述的查询权限是指查询者的查询权限，查询权限越高，其可以查到的不动产信息越多，如果查询权限为100%，则可以查看待查不动产的全部信息等。
如上述步骤S5所述，即为根据查询权限到不动产登记中心调取上述待查不动产的信息。输出方式包括多种，包括直接通过与安装有不动产信息查询系统的电脑连接的打印机打印出来；或者，通过电脑屏幕显示；或者，通过用语音播报的形式播放出来，语音播报的方式，有利于查询者是盲人时使用，相应地，上述获取查询者的身份信息的步骤，可以是通过语音识别的方式获取，如接收查询者的语音输入，然后将语音输入转换成文字等。本实施例中，如果查询上述待查不动产的信息是为了签署相关的合同，还可以将查询到的内容生成带有不动产中心公章的证明文件，以供使用。
在一具体实施例中，查询者在银行办理贷款等业务时，需要进行房产抵押等，银行工作人员可以现场准确地查看到该查询者的房产信息。具体的，银行工作人员登录不动产信息查询系统，然后使用工作人员的账户登录，在输入查询者的身份信息之后，采集查询者当前的人脸信息，并结合该查询者在公安系统的备案照片，判断该查询者是否为待查询房产的主人，若是，则开放全部的查询权限，以显示待查不动产的全部信息。另一具体实施例中，查询者在房产中介办理购房等业务时，需要查看该查询者是否满足当地的购房条件等，中介工作人员可以现场准确地查看到该查询者的房产信息，以判断该查询者是否可以购房，具体的，中介工作人员登录不动产信息查询系统，然后使用工作人员的账户登录，在输入查询者的身份信息后，采集查询者当前的人脸信息，并结合该查询者在公安系统的备案照片，判断该查询者是否为登录的身份信息的本人，如果是，则显示该查询者的全部不动产信息。进一步地，还可以链接到公积金账户或者社保账户，查看该查询者的公积金和社保缴纳等情况，进一步地，方便中介工作人员的判断。
本实施例中，上述输出与所述查询权限对应的不动产信息的步骤S5之前，包括：
S51、从所述不动产信息中抽取部分不动产信息作为真实信息；
S52、生成与所述真实信息近似的错误信息；
S53、将所述真实信息与所述错误信息生成选择题，供查询者选择；
S54、若接收到所述查询者选择的答案为所述真实信息，则执行输出与所述查询权限对应的不动产信息的步骤。
如上述步骤S51至S54所述，上述真实信息，即为待查不动产在上述不动产登记中心登记的信息，此时的待查不动产的信息是对应待查不动产在不动产登记中心登记的全部的不涉及隐私信息的信息。真实信息是从上述不动产登记中心登记的信息中抽取出来的内容，其可以随机抽取，也可以按照指定的规则抽取，如抽取不动产信息中关于转让的信息等。当获取到真实信息后，会根据语音分析方法，得到真实信息的含义，然后根据含义生成近似的错误信息，举例地，待查不动产的信息中，真实的转让记录为2010年10月15日转让，将该部分抽取出来后，生成2011年10月15日转让、2010年11月15日转让、2010年10月25日转让等干扰的错误信息，将上述的一个正确信息和三个错误信息形成“选择该不动产正确的转让记录的答案”的选择题的四个备选答案，供查询者选择，只有查询者回答正确，才会执行上述步骤S5的动作。设置回答选择题的步骤，可以进一步地判断查询者是否为待查不动产的产权人，以有效防止关于待查不动产的隐私信息被暴露。在其它实施例，可以生成多道类似的选择题，只有查询者连续答对指定数量的选择题后，才会执行上述步骤S5的动作。
在一个实施例中，上述输出与所述查询权限对应的不动产信息的步骤S5之前，包括：
S501、通过所述第一人脸照片判断所述查询者的第一年龄，以及通过所述身份信息获取所述查询者的第二年龄；
S502、判断所述第一年龄和第二年龄之间的差值是否在预设范围内；
S503、若是，则执行输出与所述查询权限对应的不动产信息的步骤。
如上述步骤S501至S503所述，上述第一年龄是不动产查询系统根据第一人脸照片推测出的年龄，具体方法是，将第一人脸照片输入到预设的年龄识别模型中，经过运算输出上述第一年龄。上述年龄识别模型是基于卷积神经网络训练出来的模型，首先获取大量的人脸照片，并在各人脸照片上标记对应的真实年龄，然后输入到预设的卷积神经网络中进行训练，训练结束后，在将另一批带有标记真实年龄的人脸照片对其进行验证，如果验证通过，即得到上述的年龄识别模型。上述第二年龄是根据身份信息计算到的年龄，如根据输入的身份证号，获取其出生日期，然后结合当前日期计算出第二年龄等。如果当前查询者是真实的产权人，那么第一年龄和第二年龄应该相同，考虑到根据模型判断的准确性问题，设置一个误差范围（上述的预设范围），只有第一年龄和第二年龄之间的差值在误差范围之内，才会认为查询者可能是待查不动产的产权人，然后才会执行上述步骤S5的动作，以有效防止关于待查不动产的隐私信息被暴露。如果第一年龄和第二年龄之间的差值在误差范围之外，可以生出“不输出查询结果的命令”，控制不动产信息查询系统停止输出查询结果，或者，生成一个更小的查询权限，以输出更少的关于待查不动产的信息。
在一个实施例中，上述计算所述第一人脸照片与所述第一备案图片的相似度值的步骤S3之前，包括：
S31、将所述第一人脸照片输入到预设的种族识别模型中，判定所述查询者的种族归属，以确定所述查询者的真实眼睛颜色；
S32、判断所述第一人脸照片中人眼的颜色是否与所述真实眼睛颜色相同；
S33、若不同，则将所述第一人脸照片中人眼的颜色还原成所述真实眼睛颜色，以便于与所述第一备案图片进行相似度比较。
如上述步骤S31至S33所述，因为查询者为了漂亮而带有美瞳，那么会影响第一人脸照片与第一备案图片相似度的计算结果，所以在此比较之前对第一人脸照片进行预处理，即将美瞳效果从第一人脸照片中清除掉。具体的过程为：先将第一人脸照片输入到种族识别模型中，判断出查询者的种族，即判断出查询者是黄种人、黑种人或白种人等，进而根据其种族确定查询者的真实眼睛颜色，如判定查询者是黄钟人，那么其眼睛应该为黑色，如果第一人脸照片中人眼颜色为其它颜色，则将其修改为黑色。上述的种族识别模型是一种具有卷积神经网络训练而得到的模型，训练方法即为常规的利用训练集进行训练，测试集进行测试等，在此不在赘述。使用预处理后的第一人脸照片，才能得到准确的第一人脸照片与第一备案图片的相似度值。
在一种可能的实施例中，上述第一人脸照片可以包括左视第一人脸照片、右视第一人脸照片和正视第一人脸照片；所述计算所述第一人脸照片与所述第一备案图片的相似度值的步骤S3，包括：
S301、将所述左视第一人脸照片、所述右视第一人脸照片和所述正视第一人脸照片分别与所述第一备案图片进行像似度比较，分别得到对应所述左视第一人脸照片、右视第一人脸照片和正视第一人脸照片的三组子相似度值；
S302、将获取到的三组子相似度值按照预设的加权平均规则计算，得到所述第一人脸照片与所述第一备案图片的相似度值。
如上述步骤S301和S302所述，上述左视第一人脸照片和右视第一人脸照片即为分别从查询者的人脸的左侧和右侧拍摄的照片。因为左视第一人脸照片和右视第一人脸照片不能涵盖全部的五官特征，所以与第一备案图片的相似度值会相对于正视第一人脸照片而言更低。本实施例中，将左视第一人脸照片、右视第一人脸照片和正视第一人脸照片分别与第一备案图片进行比较，进而得到三组子相似度值，但是并不是将三组子相似度值直接平均计算，而是加权平均计算，比如，左视第一人脸照片对应的子相似度值乘以一个百分比，右视第一人脸照片对应的子相似度值乘以一个百分比，以及正视第一人脸照片对应的子相似度值乘以一个百分比，然后相加得到最终的相似度值。本实施例拍摄查询者不同角度的人脸照片，可以防止非产权人的查询者使用产权人的照片（照片只能显示一个固定方向的影像，如一个人脸正面的照片，无法使用该正面照片拍摄出其左侧和右侧的人脸照片）欺骗不动产信息查询系统，而获取待查不动产的全部信息。
参照图2，本实施例中，上述输出与所述查询权限对应的不动产信息的步骤S5之后，包括：
S6、收集所述查询者查询其不动产的频率和地理分布；
S7、根据所述频率和所述地理分布，预测所述查询者的不动产买卖行为。
如上述步骤S7和S8所述，即为上述不动产信息查询系统会将记录的数据上传到一个服务器中，其记录的内容包括查询者的查询时间、在什么位置进行过查询等。将这些记录按照预设的方法进行预测，既可以判断出查询者的不动产买卖行为。在一具体实施例中，某查询者在三个月内，在深圳市不同的区域的房产中介查询其不动产信息，说明其在不同的区域联系过不同的中介，而中介在给其介绍房源之前，一般会核实查询者的购房条件，此时需要通过上述不动产信息查询系统查询该查询者的不动产信息，进而判断查询者是否具有购买资格。然后根据其查询的频率和其只在深圳的不同区域查找房源，既可以初步判定该查询者要买房子。而如果在三个月之内，查询者在全国大部分西区查询过其不动产信息，则说明查询者可能要在全国范围内考察房产市场，进而进行“炒房”行为。根据上述判断出的不动产买卖行为，可以给查询者发送相关的信息，以帮助查询者及时了解其关心的房产信息。
在本实施例中，上述拍摄所述查询者的第一人脸照片的步骤之后，包括：
S11、对拍摄区域内的拍摄物进行活体检测；
S12、如果检测结果为所述拍摄物为活体，则判定所述第一人脸照片为可用照片。
如上述步骤S11和S12所述，对拍摄区域内拍摄物进行活体检测，以防止他人利用不动产信息对应的产权人的照片欺骗系统。对拍摄区域内的拍摄物进行活体检测的方式包括多种，比如根据光流法，利用图像序列中的像素强度数据的时域变化和相关性来确定各自像素位置的“运动”，从图像序列中得到各个像素点的运行信息，采用高斯差分滤波器、LBP（Local Binary Pattern，是一种用来描述图像局部特征的算子）特征和支持向量机进行数据统计分析。这种活体检测方式可以在查询者无配合的情况下实现盲测。或者，利用3D摄像头拍摄人脸，得到相应的人脸区域的3D数据,并基于这些数据做进一步的分析, 最终判断出这个人脸是来自活体还是非活体。这里非活体的来源是比较广泛的,包括手机和Pad等介质的照片和视频、各种打印的不同材质的照片（这里包含各种情形的弯曲、折叠、剪裁、挖洞等情形）等。基于活体和非活体的3D人脸数据，选择最具有区分度的特征来训练分类器，利用训练好的分类器来区分活体和非活体。或者，向拍摄区域进行超声波扫描，并接收超声波的反射波；根据反射波判断拍摄区域的物体的基本轮廓；若轮廓符合预设的标准，则判定当前拍摄的图片是真人，其原理是：如果当前拍摄的图片是一个海报，那么海报必须较为平整的展开，此时其轮廓是一个平面，而如果是真人，其轮廓是一个3d轮廓等。
在一个实施例中，上述在所述预设的权限列表中查找与所述相似度值对应的查询权限的步骤S4之后，包括：收集不动产信息查询系统的每一个登录账号的每一次开放的查询权限；在指定时间长度内，若查询权限没有完全开放的次数占总查询次数的比例大于预设比值时，生成提醒信息。该提醒信息包括邮件等形式，而邮件一般发送到指定的领导邮箱。因为不动产查询是在银行或房产中介进行的，所以银行或者中介具有对应的账户，记录账户进行人脸识别过程中，查询者与备案照片不一致的次数（没有完全开放查询权限），将不一致的次数a比上识别总次数b，若比值大于预设比值，则说明该账户的工作人员在工作中认真程度不佳，在办理业务过程中不能够及时清楚地了解查询者的真实信息等，即没有沟通清楚，就进行不动产信息查询的动作，业务水平有待提高。
在其它实施例中，如果上述第一人脸照片与所述第一备案图片的相似度值达到预设值（说明第一人脸照片与第一备案图片是同一人的照片），则到网上追逃人员库中查找是否存在与查询者匹配的身份信息；如果在所述网上追逃人员库中查找到与查询者匹配的身份信息，则自动报警。因为第一人脸照片与第一备案图片是同一人的照片，所以可以确定查询者是真实有效的公民。此时，不动产信息查询系统会自动链接到网上追逃人员库中查找与查询者身份证信息相匹配的在逃人员信息，如果查找到相同的在逃人员信息，则说明查询者是网上追逃人员，系统会后台发出报警信号。报警信号可以是通知不动产信息查询系统的管理人员，比如不动产信息查询系统安装于某银行的场地，那么报警信号可以通过后台发送给银行安保部门。在另一具体实施中，还可以通过后台直接连接110报警平台，其会将对应的地址信息等相关信息发送给110报警平台等。
本实施例的不动产信息查询方法，用于在不动产登记中心之外的场所进行不动产信息线上查询，查询者无须到不动产登记中心即可查询房产信息，方便查询者到如银行、房产中介等办理需要不动产信息业务时的办理。通过人脸识别技术，进行查询者的身份验证，只有验证通过才能查询到完整的房产信息，在保护隐私的条件下，最大程度地方便了查询者；如果查询者使用虚假信息进行不动产信息登记，此时则无法查询成功，进而降低虚假登记的情况发生。
参照图3，本申请提供一种不动产信息查询装置，用于在不动产登记中心之外的场所进行不动产信息线上查询，所述装置包括：
获取单元10，用于获取查询者的身份信息；
拍摄调用单元20，用于拍摄所述查询者的第一人脸照片，以及根据所述身份信息到公安系统调用与所述身份信息对应的第一备案图片；
计算单元30，用于计算所述第一人脸照片与所述第一备案图片的相似度值；
查找单元40，用于在所述预设的权限列表中查找与所述相似度值对应的查询权限；
输出单元50，用于输出与所述查询权限对应的不动产信息。
本实施例中，上述不动产登记中心是承担不动产登记有关政策、业务、技术等的支撑工作以及国土资源法律事务工作的场所，其会在不同的行政区域设置对应的办事处等，比如深圳市不动产登记中心，其主要办理深圳市的不动产登记业务等。上述不动产登记中心之外的场所进行不动产信息线上查询，主要是指被不动产登记中心授权的银行、房产中介等企业的办公场所，其使用的电脑等电子设备上安装有上述不动产信息查询系统，不动产信息查询系统是通过http协议的接口与不动产登记中心的服务器连接。本实施例中，只有授权的企业的电脑等电子设备才可以登录使用上述不动产信息查询系统。
在获取单元10中，一般在办公设备上安装不动产登记业务系统，上述不动产查询系统呈现在计算机的屏幕上的内容是一个线上查询页面，其设置有输入查询者身份信息的输入窗口等。其中，身份信息至少包括查询者的身份证号。本申请中，上述身份信息一般是由银行或房产中介的工作人员手动输入到不动产查询系统中。在其它实施例中，还可以通过输入待查不动产的基本信息反向推导出查询者的身份信息，即根据待查不动产的基本信息到不动产中心查找与其对应的产权人信息，该产权人信息即为默认的查询者信息。
在上述拍摄调用单元20中，上述不动产信息查询系统会控制与其连接的摄像头拍摄查询者的脸部照片（通常包括正脸照），即上述的第一人脸照片。上述第一备案图片是指查询者办理身份证时的图片。本实施例中，上述不动产信息查询系统是公安系统授权的系统，公安系统允许不动产信息查询系统访问其存储有身份证的数据库，因此不动产信息查询系统可以根据获取到的身份证号到公安系统中调用与所述身份信息对应的身份证图片，而身份证图片上有人脸图片，该人脸图片即为上述的第一备案图片。
在上述计算单元30中，相似度值的计算过程具体为：将第一人脸照片输入到预设的人脸识别模型中进行运算输出第一人脸照片的特征向量，以及将第一备案图片输入到上述的人脸识别模型中进行运算输出第一备案图片的特征向量，然后将第一人脸照片的特征向量和第一备案图片的特征向量进行相似度计算。具体的相似度值计算可以使用欧几里得距离（Eucledian Distance）、曼哈顿距离（Manhattan Distance）、明可夫斯基距离（Minkowski distance）或者余弦相似度中一种算法进行计算。
在上述查找单元40中，上述预设的权限列表是一个一对一的映射列表，其中有相似度值的一列，以及查询权限的一列，每一个相似度值对应一个查询权限。在本实施例中，当在权限列表无法查找到与上述计算单元30计算出相似度值相同的相似度值时，则选择权限列表中与上述刘表计算出相似度值差值最小的相似度值使用。比如，计算出的第一人脸照片与所述第一备案图片的相似度值为10，而权限列表中没有相似度值为10的选项，而只有3、6、9、12等相似度值选项，那么选择与相似度值10接近的相似度值9，然后选择相似度值9对应的查询权限。上述的查询权限是指查询者的查询权限，查询权限越高，其可以查到的不动产信息越多，如果查询权限为100%，则可以查看待查不动产的全部信息等。
在上述输出单元50中，即为根据查询权限到不动产登记中心调取上述待查不动产的信息的单元。输出方式包括多种，包括直接通过与安装有不动产信息查询系统的电脑连接的打印机打印出来；或者，通过电脑屏幕显示；或者，通过用语音播报的形式播放出来，语音播报的方式，有利于查询者是盲人时使用，相应地，上述获取查询者的身份信息的方法，可以是通过语音识别的方式获取，如接收查询者的语音输入，然后将语音输入转换成文字等。本实施例中，如果查询上述待查不动产的信息是为了签署相关的合同，还可以将查询到的内容生成带有不动产中心公章的证明文件，以供使用。
在一具体实施例中，查询者在银行办理贷款等业务时，需要进行房产抵押等，银行工作人员可以现场准确地查看到该查询者的房产信息。具体的，银行工作人员登录不动产信息查询系统，然后使用工作人员的账户登录，在输入查询者的身份信息之后，采集查询者当前的人脸信息，并结合该查询者在公安系统的备案照片，判断该查询者是否为待查询房产的主人，若是，则开放全部的查询权限，以显示待查不动产的全部信息。另一具体实施例中，查询者在房产中介办理购房等业务时，需要查看该查询者是否满足当地的购房条件等，中介工作人员可以现场准确地查看到该查询者的房产信息，以判断该查询者是否可以购房，具体的，中介工作人员登录不动产信息查询系统，然后使用工作人员的账户登录，在输入查询者的身份信息后，采集查询者当前的人脸信息，并结合该查询者在公安系统的备案照片，判断该查询者是否为登录的身份信息的本人，如果是，则显示该查询者的全部不动产信息。进一步地，还可以链接到公积金账户或者社保账户，查看该查询者的公积金和社保缴纳等情况，进一步地，方便中介工作人员的判断。
参照图4，本实施例中，上述不动产信息查询装置，包括：
抽取单元51，用于从所述不动产信息中抽取部分不动产信息作为真实信息；
生成单元52，用于生成与所述真实信息近似的错误信息；
混合单元53，用于将所述真实信息与所述错误信息生成选择题，供查询者选择；
接收生成单元54，用于若接收到所述查询者选择的答案，则执行输出与所述查询权限对应的不动产信息的步骤。
在上述抽取单元51、生成单元52、混合单元53和接收生成单元54中，上述不动产的真实信息，即为待查不动产在上述不动产登记中心登记的信息，此时的待查不动产的信息是对应待查不动产在不动产登记中心登记的全部的不涉及隐私信息的信息。真实信息是从上述不动产登记中心登记的信息中抽取出来的内容，其可以随机抽取，也可以按照指定的规则抽取，如抽取不动产信息中关于转让的信息等。当获取到真实信息后，会根据语音分析方法，得到真实信息的含义，然后根据含义生成近似的错误信息，举例地，待查不动产的信息中，真实的转让记录为2010年10月15日转让，将该部分抽取出来后，生成2011年10月15日转让、2010年11月15日转让、2010年10月25日转让等干扰的错误信息，将上述的一个正确信息和三个错误信息形成“选择该不动产正确的转让记录的答案”的选择题的四个备选答案，供查询者选择，只有查询者回答正确，才会输出对应查询权限的不动产信息。设置回答选择题的步骤，可以进一步地判断查询者是否为待查不动产的产权人，以有效防止关于待查不动产的隐私信息被暴露。在其它实施例，可以生成多道类似的选择题，只有查询者连续答对指定数量的选择题后，才会输出对应查询权限的不动产信息。
参照图5，在一个实施例中，上述不动产信息查询装置，包括：
年龄获取单元501，用于通过所述第一人脸照片判断所述查询者的第一年龄，以及通过所述身份信息获取所述查询者的第二年龄；
判断单元502，用于判断所述第一年龄和第二年龄之间的差值是否在预设范围内；
生成指令单元503，用于若第一年龄和第二年龄之间的差值在预设范围内，则执行输出与所述查询权限对应的不动产信息的步骤。
在上述年龄获取单元501、判断单元502和生成指令单元503中，上述第一年龄是不动产查询系统根据第一人脸照片推测出的年龄，具体方法是，将第一人脸照片输入到预设的年龄识别模型中，经过运算输出上述第一年龄。上述年龄识别模型是基于卷积神经网络训练出来的模型，首先获取大量的人脸照片，并在各人脸照片上标记对应的真实年龄，然后输入到预设的卷积神经网络中进行训练，训练结束后，在将另一批带有标记真实年龄的人脸照片对其进行验证，如果验证通过，即得到上述的年龄识别模型。上述第二年龄是根据身份信息计算到的年龄，如根据输入的身份证号，获取其出生日期，然后结合当前日期计算出第二年龄等。如果当前查询者是真实的产权人，那么第一年龄和第二年龄应该相同，考虑到根据模型判断的准确性问题，设置一个误差范围（上述的预设范围），只有第一年龄和第二年龄之间的差值在误差范围之内，才会认为查询者可能是待查不动产的产权人，然后才会输出对应查询权限的不动产信息，以有效防止关于待查不动产的隐私信息被暴露。如果第一年龄和第二年龄之间的差值在误差范围之外，可以生出“不输出查询结果的命令”，控制不动产信息查询系统停止输出查询结果，或者，生成一个更小的查询权限，以输出更少的关于待查不动产的信息。
参照图6，在一个实施例中，上述不动产信息查询装置，包括：
判定单元31，用于将所述第一人脸照片输入到预设的种族识别模型中，判定所述查询者的种族归属，以确定所述查询者的真实眼睛颜色；
颜色判断单元32，用于判断所述第一人脸照片中人眼的颜色是否与所述真实眼睛颜色相同；
还原单元33，用于若第一人脸照片中人眼的颜色与所述真实眼睛颜色不相同，则将所述第一人脸照片中人眼的颜色还原成所述真实眼睛颜色，以便于与所述第一备案图片进行相似度比较。
在上述判定单元31、颜色判断单元32和还原单元33中，因为查询者为了漂亮而带有美瞳，那么会影响第一人脸照片与第一备案图片相似度的计算结果，所以在此比较之前对第一人脸照片进行预处理，即将美瞳效果从第一人脸照片中清除掉。具体的过程为：先将第一人脸照片输入到种族识别模型中，判断出查询者的种族，即判断出查询者是黄种人、黑种人或白种人等，进而根据其种族确定查询者的真实眼睛颜色，如判定查询者是黄钟人，那么其眼睛应该为黑色，如果第一人脸照片中人眼颜色为其它颜色，则将其修改为黑色。上述的种族识别模型是一种具有卷积神经网络训练而得到的模型，训练方法即为常规的利用训练集进行训练，测试集进行测试等，在此不在赘述。使用预处理后的第一人脸照片，才能得到准确的第一人脸照片与第一备案图片的相似度值。
参照图7，在一种可能的实施例中，上述第一人脸照片可以包括左视第一人脸照片、右视第一人脸照片和正视第一人脸照片；上述计算单元30，包括：
分别比较模块301，用于将所述左视第一人脸照片、所述右视第一人脸照片和所述正视第一人脸照片分别与所述第一备案图片进行像似度比较，分别得到对应所述左视第一人脸照片、右视第一人脸照片和正视第一人脸照片的三组子相似度值；
加权计算模块302，用于将获取到的三组子相似度值按照预设的加权平均规则计算，得到所述第一人脸照片与所述第一备案图片的相似度值。
在上述分别比较模块301和加权计算模块302中，上述左视第一人脸照片和右视第一人脸照片即为分别从查询者的人脸的左侧和右侧拍摄的照片。因为左视第一人脸照片和右视第一人脸照片不能涵盖全部的五官特征，所以与第一备案图片的相似度值会相对于正视第一人脸照片而言更低。本实施例中，将左视第一人脸照片、右视第一人脸照片和正视第一人脸照片分别与第一备案图片进行比较，进而得到三组子相似度值，但是并不是将三组子相似度值直接平均计算，而是加权平均计算，比如，左视第一人脸照片对应的子相似度值乘以一个百分比，右视第一人脸照片对应的子相似度值乘以一个百分比，以及正视第一人脸照片对应的子相似度值乘以一个百分比，然后相加得到最终的相似度值。本实施例拍摄查询者不同角度的人脸照片，可以防止非产权人的查询者使用产权人的照片（照片只能显示一个固定方向的影像，如一个人脸正面的照片，无法使用该正面照片拍摄出其左侧和右侧的人脸照片）欺骗不动产信息查询系统，而获取待查不动产的全部信息。
参照图8，本实施例中，上述不动产信息查询装置，还包括：
收集单元60，用于收集所述查询者查询其不动产的频率和地理分布；
预测单元70，用于根据所述频率和所述地理分布，预测所述查询者的不动产买卖行为。
在上述收集单元60和预测单元70中，上述不动产信息查询系统会将记录的数据上传到一个服务器中，其记录的内容包括查询者的查询时间、在什么位置进行过查询等。将这些记录按照预设的方法进行预测，既可以判断出查询者的不动产买卖行为。在一具体实施例中，某查询者在三个月内，在深圳市不同的区域的房产中介查询其不动产信息，说明其在不同的区域联系过不同的中介，而中介在给其介绍房源之前，一般会核实查询者的购房条件，此时需要通过上述不动产信息查询系统查询该查询者的不动产信息，进而判断查询者是否具有购买资格。然后根据其查询的频率和其只在深圳的不同区域查找房源，既可以初步判定该查询者要买房子。而如果在三个月之内，查询者在全国大部分西区查询过其不动产信息，则说明查询者可能要在全国范围内考察房产市场，进而进行“炒房”行为。根据上述判断出的不动产买卖行为，可以给查询者发送相关的信息，以帮助查询者及时了解其关心的房产信息。
在本实施例中，上述不动产信息查询装置，还包括：
活体检测单元，用于对拍摄区域内的拍摄物进行活体检测；
活体判定单元，用于如果检测结果为所述拍摄物为活体，则判定所述第一人脸照片为可用照片。
对拍摄区域内拍摄物进行活体检测，以防止他人利用不动产信息对应的产权人的照片欺骗系统。对拍摄区域内的拍摄物进行活体检测的方式包括多种，比如根据光流法，利用图像序列中的像素强度数据的时域变化和相关性来确定各自像素位置的“运动”，从图像序列中得到各个像素点的运行信息，采用高斯差分滤波器、LBP（Local Binary Pattern，是一种用来描述图像局部特征的算子）特征和支持向量机进行数据统计分析。这种活体检测方式可以在查询者无配合的情况下实现盲测。或者，利用3D摄像头拍摄人脸，得到相应的人脸区域的3D数据,并基于这些数据做进一步的分析, 最终判断出这个人脸是来自活体还是非活体。这里非活体的来源是比较广泛的,包括手机和Pad等介质的照片和视频、各种打印的不同材质的照片（这里包含各种情形的弯曲、折叠、剪裁、挖洞等情形）等。基于活体和非活体的3D人脸数据，选择最具有区分度的特征来训练分类器，利用训练好的分类器来区分活体和非活体。或者，向拍摄区域进行超声波扫描，并接收超声波的反射波；根据反射波判断拍摄区域的物体的基本轮廓；若轮廓符合预设的标准，则判定当前拍摄的图片是真人，其原理是：如果当前拍摄的图片是一个海报，那么海报必须较为平整的展开，此时其轮廓是一个平面，而如果是真人，其轮廓是一个3d轮廓等。
在一个实施例中，上述不动产信息查询装置，还包括：收集权限单元，用于收集不动产信息查询系统的每一个登录账号的每一次开放的查询权限；提醒单元，用于在指定时间长度内，若查询权限没有完全开放的次数占总查询次数的比例大于预设比值时，生成提醒信息。该提醒信息包括邮件等形式，而邮件一般发送到指定的领导邮箱。因为不动产查询是在银行或房产中介进行的，所以银行或者中介具有对应的账户，记录账户进行人脸识别过程中，查询者与备案照片不一致的次数（没有完全开放查询权限），将不一致的次数a比上识别总次数b，若比值大于预设比值，则说明该账户的工作人员在工作中认真程度不佳，在办理业务过程中不能够及时清楚地了解查询者的真实信息等，即没有沟通清楚，就进行不动产信息查询的动作，业务水平有待提高。
在其它实施例中，如果上述第一人脸照片与所述第一备案图片的相似度值达到预设值（说明第一人脸照片与第一备案图片是同一人的照片），则到网上追逃人员库中查找是否存在与查询者匹配的身份信息；如果在所述网上追逃人员库中查找到与查询者匹配的身份信息，则自动报警。因为第一人脸照片与第一备案图片是同一人的照片，所以可以确定查询者是真实有效的公民。此时，不动产信息查询系统会自动链接到网上追逃人员库中查找与查询者身份证信息相匹配的在逃人员信息，如果查找到相同的在逃人员信息，则说明查询者是网上追逃人员，系统会后台发出报警信号。报警信号可以是通知不动产信息查询系统的管理人员，比如不动产信息查询系统安装于某银行的场地，那么报警信号可以通过后台发送给银行安保部门。在另一具体实施中，还可以通过后台直接连接110报警平台，其会将对应的地址信息等相关信息发送给110报警平台等。
本实施例的不动产信息查询装置，用于在不动产登记中心之外的场所进行不动产信息线上查询，查询者无须到不动产登记中心即可查询房产信息，方便查询者到如银行、房产中介等办理需要不动产信息业务时的办理。通过人脸识别技术，进行查询者的身份验证，只有验证通过才能查询到完整的房产信息，在保护隐私的条件下，最大程度地方便了查询者；如果查询者使用虚假信息进行不动产信息登记，此时则无法查询成功，进而降低虚假登记的情况发生。
参照图9，本申请实施例中还提供一种计算机设备，该计算机设备可以是服务器，其内部结构可以如图9所示。该计算机设备包括通过系统总线连接的处理器、存储器、网络接口和数据库。其中，该计算机设计的处理器用于提供计算和控制能力。该计算机设备的存储器包括非易失性存储介质、内存储器。该非易失性存储介质存储有操作系统、计算机可读指令和数据库。该内存器为非易失性存储介质中的操作系统和计算机可读指令的运行提供环境。该计算机设备的数据库用于存储人脸识别模型等数据。该计算机设备的网络接口用于与外部的终端通过网络连接通信。该计算机可读指令被处理器执行时以实现上述各方法实施例的流程。
本申请实施例还提供一种计算机非易失性可读存储介质，其上存储有计算机可读指令，计算机可读指令被处理器执行时实现上述各方法实施例的流程。
以上所述仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本申请的专利保护范围内。

Claims (20)

  1. 一种不动产信息查询方法，其特征在于，用于在不动产登记中心之外的场所进行不动产信息线上查询，所述方法包括：
    获取查询者的身份信息；
    拍摄所述查询者的第一人脸照片，以及根据所述身份信息到公安系统调用与所述身份信息对应的第一备案图片；
    计算所述第一人脸照片与所述第一备案图片的相似度值；
    在所述预设的权限列表中查找与所述相似度值对应的查询权限；
    输出与所述查询权限对应的不动产信息。
  2. 根据权利要求1所述的不动产信息查询方法，其特征在于，所述输出与所述查询权限对应的不动产信息的步骤之前，包括：
    从所述不动产信息中抽取部分不动产信息作为真实信息；
    生成与所述真实信息近似的错误信息；
    根据所述真实信息与所述错误信息生成选择题，供查询者选择；
    若接收到所述查询者选择的答案为所述真实信息，则执行输出与所述查询权限对应的不动产信息的步骤。
  3. 根据权利要求1所述的不动产信息查询方法，其特征在于，所述输出与所述查询权限对应的不动产信息的步骤之前，包括：
    通过所述第一人脸照片判断所述查询者的第一年龄，以及通过所述身份
    信息获取所述查询者的第二年龄；
    判断所述第一年龄和第二年龄之间的差值是否在预设范围内；
    若是，则执行输出与所述查询权限对应的不动产信息的步骤。
  4. 根据权利要求1所述的不动产信息查询方法，其特征在于，所述计算所述第一人脸照片与所述第一备案图片的相似度值的步骤之前，包括：
    将所述第一人脸照片输入到预设的种族识别模型中，判定所述查询者的种族归属，以确定所述查询者的真实眼睛颜色；
    判断所述第一人脸照片中人眼的颜色是否与所述真实眼睛颜色相同；
    若不同，则将所述第一人脸照片中人眼的颜色还原成所述真实眼睛颜色，以便于与所述第一备案图片进行相似度比较。
  5. 根据权利要求1所述的不动产信息查询方法，其特征在于，所述第一人脸照片包括左视第一人脸照片、右视第一人脸照片和正视第一人脸照片；所述计算所述第一人脸照片与所述第一备案图片的相似度值的步骤包括：
    将所述左视第一人脸照片、所述右视第一人脸照片和所述正视第一人脸照片分别与所述第一备案图片进行像似度比较，分别得到对应所述左视第一人脸照片、右视第一人脸照片和正视第一人脸照片的三组子相似度值；
    将获取到的三组子相似度值按照预设的加权平均规则计算，得到所述第一人脸照片与所述第一备案图片的相似度值。
  6. 根据权利要求1所述的不动产信息查询方法，其特征在于，所述方法还包括：
    收集所述查询者查询其不动产的频率和地理分布；
    根据所述频率和所述地理分布，预测所述查询者的不动产买卖行为。
  7. 根据权利要求1所述的不动产信息查询方法，其特征在于，所述拍摄所述查询者的第一人脸照片的步骤之后，包括：
    对拍摄区域内的拍摄物进行活体检测；
    如果检测结果为所述拍摄物为活体，则判定所述第一人脸照片为可用照片。
  8. 一种不动产信息查询装置，其特征在于，用于在不动产登记中心之外的场所进行不动产信息线上查询，所述装置包括：
    获取单元，用于不动产信息查询系统获取查询者的身份信息；
    拍摄调用单元，用于拍摄所述查询者的第一人脸照片，以及根据所述身份信息到公安系统调用与所述身份信息对应的第一备案图片；
    计算单元，用于计算所述第一人脸照片与所述第一备案图片的相似度值；
    查找单元，用于在所述预设的权限列表中查找与所述相似度值对应的查询权限；
    输出单元，用于输出与所述查询权限对应的不动产信息。
  9. 根据权利要求8所述的不动产信息查询装置，其特征在于，所述不动产信息查询装置，包括：
    抽取单元，用于从所述不动产信息中抽取部分不动产信息作为真实信息；
    生成单元，用于生成与所述真实信息近似的错误信息；
    混合单元，用于将所述真实信息与所述错误信息生成选择题，供查询者选择；
    接收生成单元，用于若接收到所述查询者选择的答案，则执行输出与所述查询权限对应的不动产信息的步骤。
  10. 根据权利要求8所述的不动产信息查询装置，其特征在于，所述不动产信息查询装置，包括：
    年龄获取单元，用于通过所述第一人脸照片判断所述查询者的第一年龄，以及通过所述身份信息获取所述查询者的第二年龄；
    判断单元，用于判断所述第一年龄和第二年龄之间的差值是否在预设范围内；
    生成指令单元，用于若第一年龄和第二年龄之间的差值在预设范围内，则执行输出与所述查询权限对应的不动产信息的步骤。
  11. 根据权利要求8所述的不动产信息查询装置，其特征在于，所述不动产信息查询装置，包括：
    判定单元，用于将所述第一人脸照片输入到预设的种族识别模型中，判定所述查询者的种族归属，以确定所述查询者的真实眼睛颜色；
    颜色判断单元，用于判断所述第一人脸照片中人眼的颜色是否与所述真实眼睛颜色相同；
    还原单元，用于若第一人脸照片中人眼的颜色与所述真实眼睛颜色不相同，则将所述第一人脸照片中人眼的颜色还原成所述真实眼睛颜色，以便于与所述第一备案图片进行相似度比较。
  12. 根据权利要8所述的不动产信息查询装置，其特征在于，所述第一人脸照片包括左视第一人脸照片、右视第一人脸照片和正视第一人脸照片；所述计算单元，包括：
    分别比较模块，用于将所述左视第一人脸照片、所述右视第一人脸照片和所述正视第一人脸照片分别与所述第一备案图片进行像似度比较，分别得到对应所述左视第一人脸照片、右视第一人脸照片和正视第一人脸照片的三组子相似度值；
    加权计算模块，用于将获取到的三组子相似度值按照预设的加权平均规则计算，得到所述第一人脸照片与所述第一备案图片的相似度值。
  13. 根据权利要求8所述的不动产信息查询装置，其特征在于，所述不动产信息查询装置，还包括：
    收集单元，用于收集所述查询者查询其不动产的频率和地理分布；
    预测单元，用于根据所述频率和所述地理分布，预测所述查询者的不动产买卖行为。
  14. 根据权利要求8所述的不动产信息查询装置，其特征在于，所述不动产信息查询装置，还包括：
    活体检测单元，用于对拍摄区域内的拍摄物进行活体检测；
    活体判定单元，用于如果检测结果为所述拍摄物为活体，则判定所述第一人脸照片为可用照片。
  15. 一种计算机设备，包括存储器和处理器，所述存储器存储有计算机可读指令，其特征在于，所述处理器执行所述计算机可读指令时实现不动产信息查询方法，用于在不动产登记中心之外的场所进行不动产信息线上查询，所述方法包括：
    获取查询者的身份信息；
    拍摄所述查询者的第一人脸照片，以及根据所述身份信息到公安系统调用与所述身份信息对应的第一备案图片；
    计算所述第一人脸照片与所述第一备案图片的相似度值；
    在所述预设的权限列表中查找与所述相似度值对应的查询权限；
    输出与所述查询权限对应的不动产信息。
  16. 根据权利要求15所述的计算机设备，其特征在于，所述输出与所述查询权限对应的不动产信息的步骤之前，包括：
    从所述不动产信息中抽取部分不动产信息作为真实信息；
    生成与所述真实信息近似的错误信息；
    根据所述真实信息与所述错误信息生成选择题，供查询者选择；
    若接收到所述查询者选择的答案为所述真实信息，则执行输出与所述查询权限对应的不动产信息的步骤。
  17. 根据权利要求15所述的计算机设备，其特征在于，所述输出与所述查询权限对应的不动产信息的步骤之前，包括：
    通过所述第一人脸照片判断所述查询者的第一年龄，以及通过所述身份
    信息获取所述查询者的第二年龄；
    判断所述第一年龄和第二年龄之间的差值是否在预设范围内；
    若是，则执行输出与所述查询权限对应的不动产信息的步骤。
  18. 根据权利要求15所述的计算机设备，其特征在于，所述计算所述第一人脸照片与所述第一备案图片的相似度值的步骤之前，包括：
    将所述第一人脸照片输入到预设的种族识别模型中，判定所述查询者的种族归属，以确定所述查询者的真实眼睛颜色；
    判断所述第一人脸照片中人眼的颜色是否与所述真实眼睛颜色相同；
    若不同，则将所述第一人脸照片中人眼的颜色还原成所述真实眼睛颜色，以便于与所述第一备案图片进行相似度比较。
  19. 一种计算机非易失性可读存储介质，其上存储有计算机可读指令，其特征在于，所述计算机可读指令被处理器执行时实现不动产信息查询方法，用于在不动产登记中心之外的场所进行不动产信息线上查询，所述方法包括：
    获取查询者的身份信息；
    拍摄所述查询者的第一人脸照片，以及根据所述身份信息到公安系统调用与所述身份信息对应的第一备案图片；
    计算所述第一人脸照片与所述第一备案图片的相似度值；
    在所述预设的权限列表中查找与所述相似度值对应的查询权限；
    输出与所述查询权限对应的不动产信息。
  20. 根据权利要求19所述的计算机非易失性可读存储介质，其特征在于，所述输出与所述查询权限对应的不动产信息的步骤之前，包括：
    从所述不动产信息中抽取部分不动产信息作为真实信息；
    生成与所述真实信息近似的错误信息；
    根据所述真实信息与所述错误信息生成选择题，供查询者选择；
    若接收到所述查询者选择的答案为所述真实信息，则执行输出与所述查询权限对应的不动产信息的步骤。
PCT/CN2018/095387 2018-05-04 2018-07-12 不动产信息查询方法、装置、计算机设备和存储介质 WO2019210575A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810422089.9A CN108846015A (zh) 2018-05-04 2018-05-04 不动产信息查询方法、装置、计算机设备和存储介质
CN201810422089.9 2018-05-04

Publications (1)

Publication Number Publication Date
WO2019210575A1 true WO2019210575A1 (zh) 2019-11-07

Family

ID=64212802

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/095387 WO2019210575A1 (zh) 2018-05-04 2018-07-12 不动产信息查询方法、装置、计算机设备和存储介质

Country Status (2)

Country Link
CN (1) CN108846015A (zh)
WO (1) WO2019210575A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111383032A (zh) * 2020-02-12 2020-07-07 北京城市网邻信息技术有限公司 一种房源信息的真实性检测方法和装置

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109636318B (zh) * 2018-11-23 2020-01-24 南宁市国土测绘地理信息中心 一种不动产登记系统及不动产登记的方法
CN109656962A (zh) * 2018-12-12 2019-04-19 南宁市国土测绘地理信息中心 一种不动产网络司法查控平台
CN111611418A (zh) * 2019-02-25 2020-09-01 阿里巴巴集团控股有限公司 数据存储方法及数据查询方法
TWI739372B (zh) * 2020-04-07 2021-09-11 中華誠信資產管理顧問股份有限公司 不動產比較價格地理資訊評估方法及其系統
CN112667718B (zh) * 2020-12-24 2021-07-16 广西中科曙光云计算有限公司 一种不动产信息的共享查询方法及系统

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1779704A (zh) * 2004-11-18 2006-05-31 黄金富 信用资料查询系统及查询方法
US9032490B1 (en) * 2012-09-12 2015-05-12 Emc Corporation Techniques for authenticating a user with heightened security
CN107315782A (zh) * 2017-06-08 2017-11-03 北京奇艺世纪科技有限公司 一种数据查询方法及装置
CN107680294A (zh) * 2017-09-11 2018-02-09 平安科技（深圳）有限公司 房产信息查询方法、系统、终端设备及存储介质

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106203294B (zh) * 2016-06-30 2019-05-03 广东微模式软件股份有限公司 基于人脸属性分析的人证合一身份验证方法
CN106981140A (zh) * 2017-03-30 2017-07-25 广东微模式软件股份有限公司 一种电话卡自助服务一体化装置及其方法
CN107301551B (zh) * 2017-07-15 2020-09-29 刘兴丹 一种网络支付前搜索、查询、验证的方法、装置、系统

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1779704A (zh) * 2004-11-18 2006-05-31 黄金富 信用资料查询系统及查询方法
US9032490B1 (en) * 2012-09-12 2015-05-12 Emc Corporation Techniques for authenticating a user with heightened security
CN107315782A (zh) * 2017-06-08 2017-11-03 北京奇艺世纪科技有限公司 一种数据查询方法及装置
CN107680294A (zh) * 2017-09-11 2018-02-09 平安科技（深圳）有限公司 房产信息查询方法、系统、终端设备及存储介质

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111383032A (zh) * 2020-02-12 2020-07-07 北京城市网邻信息技术有限公司 一种房源信息的真实性检测方法和装置
CN111383032B (zh) * 2020-02-12 2023-11-14 北京城市网邻信息技术有限公司 一种房源信息的真实性检测方法和装置

Also Published As

Publication number Publication date
CN108846015A (zh) 2018-11-20

Similar Documents

Publication Publication Date Title
WO2019210575A1 (zh) 不动产信息查询方法、装置、计算机设备和存储介质
CN109636318B (zh) 一种不动产登记系统及不动产登记的方法
CN112418009B (zh) 一种图像质量检测方法、终端设备及存储介质
AU2018201140B2 (en) System and method for candidate profile screening
WO2021190086A1 (zh) 面审风险控制方法、装置、计算机设备及存储介质
US20200294130A1 (en) Loan matching system and method
CN107025397B (zh) 身份信息的获取方法和装置
WO2019200737A1 (zh) 不动产资料上传方法、装置、计算机设备和存储介质
TW202026948A (zh) 活體檢測方法、裝置以及儲存介質
CN111886842A (zh) 使用基于阈值的匹配进行远程用户身份验证
CN113014566B (zh) 恶意注册的检测方法、装置、计算机可读介质及电子设备
CN111325082A (zh) 一种人员专注度分析方法及装置
US20210224588A1 (en) Recruitment process graph based unsupervised anomaly detection
Barrington et al. A comparative analysis of human and AI performance in forensic estimation of physical attributes
KR20200087036A (ko) 블록체인 기반의 기업 직무정보 수집방법 및 서버
Perritt Jr Defending Face-Recognition Technology (And Defending Against It)
CN115147705B (zh) 人脸翻拍检测方法、装置、电子设备及存储介质
EP4064119A1 (en) Method and system for identifying a person
CN112287830A (zh) 一种图像的检测方法及装置
CN115438888A (zh) 交互服务质量检测方法、电子设备及计算机存储介质
CN111767845A (zh) 证件识别方法及装置
CN117764710B (zh) 一种住房金融风险行为的监控方法
KR102479930B1 (ko) 부동산 정보 수집 방법 및 그 장치
Nguyen et al. PROCTORING ONLINE EXAM USING IOT TECHNOLOGY
CN105975636A (zh) 一种优化在线咨询服务的方法和装置

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: 18917301

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: 18917301

Country of ref document: EP

Kind code of ref document: A1