A method for improving the quality of real estate integrated data based on multi-source data

Abstract

The present invention relates to the field of real estate data technology, and in particular to a real estate data quality improvement method based on multi-source data, which comprises obtaining a real-life picture of a house according to a shooting device, reading existing data in a database, and constructing a three-dimensional model of a house according to a topographic map and a real-life picture of the house in the existing data, obtaining real estate data, and obtaining a cadastral map, a natural building of real estate and a plot map of real estate in the real estate data, constructing a database by comparing and analyzing the three-dimensional model of the house with the real estate data, and outputting a comparative analysis report, updating data in an original database of real estate through superposition analysis of a topographic map and a cadastral map, so that the data quality of the original database is higher, and outputting a more accurate comparative analysis report for cadastral information and plot information in the real estate, through area comparison between a topographic map and a cadastral map, so that subsequent planning work is more convenient.

Description

Technical Field

[0001] The present invention relates to the field of real estate integration data technology, and in particular to a real estate integration data quality improvement method based on multi-source data. Background Art

[0002] The principle of integration of real estate and land means that the ownership of land use rights and buildings and fixtures on the land, such as houses, belongs to the same entity. When real estate is transferred or mortgaged, the ownership of the house and the land must be transferred or mortgaged at the same time. The confirmation of rural real estate and land rights is to implement unified registration procedures for the use rights of rural homesteads and collective construction land and the buildings and structures on the land. However, most houses now have certain incomplete information, resulting in incomplete real estate and land information or inaccurate real estate and land data. In the existing database, there are also errors in the three-dimensional modeling of houses and the natural building lines of houses to determine the ownership of land use rights and buildings and fixtures on the land, and the real estate and land system data is redundant. Therefore, updating the real estate and land system data involves factors of large computing power and high cost. A real estate and land data quality improvement method is urgently needed to improve the real estate and land system data quality, and make the real estate and land system data more accurate while making the real estate and land system data lightweight. Summary of the invention

[0003] The purpose of the present invention is to propose a method for improving the quality of integrated real estate data based on multi-source data to solve one or more technical problems existing in the prior art and at least provide a beneficial choice or create conditions.

[0004] A method for improving the quality of real estate integrated data based on multi-source data, the method comprising the following steps:

[0005] S100: Acquire a real-life picture of the house using a shooting device;

[0006] S200: Reading existing data in the database and constructing a three-dimensional model of the house according to the topographic map and the real-life picture of the house in the existing data;

[0007] S300: Acquire the real estate data, and acquire the cadastral map, the real estate natural building and the real estate plot map from the real estate data;

[0008] S400: Building a database by combining the three-dimensional model of the house with the real estate integrated data for comparative analysis, and outputting a comparative analysis report.

[0009] Furthermore, in step S100 and step S200, a real-life picture of the house is taken by a drone, the real-life picture of the house including: a roof picture and a side picture of the house, a three-dimensional model of the house is constructed by combining the house topographic map in a database with oblique photography, the roof picture of the house taken by the drone is mapped onto the top of the three-dimensional model of the house, and the roof picture of the house is combined with the topographic map taken by the drone to divide the roof of the house into a topographic top layer and a secondary top layer, and the elevations of various parts of the roof in the topographic top layer and the secondary top layer are determined, the building boundary line is determined according to the roof picture of the house, and a topographic roof picture of the house is obtained by comparing and analyzing the three-dimensional model of the house with the house topographic map, the topographic roof picture of the house being divided into a topographic top layer and a secondary top layer.

[0010] Furthermore, in step S300, a cadastral map, a natural building of integrated real estate and land, and a plot map of integrated real estate and land are obtained through a database, wherein the cadastral map indicates the ownership of the land in the topographic map, the natural building of integrated real estate and land indicates the exact location of the house in the integrated real estate and land system, and the plot map of integrated real estate and land reflects the basic situation of the plot, which includes: the plot ownership boundary, the location of boundary points, the location and nature of the buildings in the plot, and the baseline of the natural building of integrated real estate and land is determined according to the natural building of integrated real estate and land, and the partition ownership of the natural building is determined according to the baseline of the natural building.

[0011] Further, in step S400, the cadastral map part in the cadastral map is divided into a cadastral top layer and a cadastral second top layer, and the cadastral top layer and the cadastral second top layer are spatially superimposed and compared with the house topography top layer and the house topography second top layer, respectively. The specific method of superimposed and compared analysis includes the following steps:

[0012] S401: compare the top layer of the house terrain with the edge of the top layer of the cadastral map, and correct the edge of the top layer of the cadastral map by giving a higher weight to the top layer of the house terrain, so that the area of ​​the top layer of the cadastral map is recalculated, and the number of pixel points covered by the edge line is compared with the direction of the edge line through the edge line comparison to determine the correct direction of the edge line, and unify the cadastral information of the top layer sky map based on the correct direction of the edge line;

[0013] S402: Obtain the topographic sub-top floor of the house and the cadastral sub-top floor, determine the building range line of the edge part in the topographic sub-top floor of the house, re-determine the outer edge baseline of the cadastral sub-top floor according to the building range line, and determine the outer edge baseline according to the cadastral information of the top floor determined in step S401, determine the cadastral map top floor according to the outer edge baseline and the outer edge baseline, and perform weight matching on the area of ​​the cadastral sub-top floor determined according to the outer edge baseline and the inner edge baseline and the original cadastral sub-top floor area;

[0014] S403: The step of weighting the cadastral map area is as follows: record the area of ​​the cadastral second top floor determined by the external edge baseline and the internal edge baseline as CD1, define the original cadastral second top floor area as CD, the original cadastral second top floor area is CD, obtain the weight ratio of the original cadastral second top floor area as CD by comparing the comparison ratio of the topographic top floor and the cadastral top floor, and obtain the top floor area data difference in the building. The sequence Ck is constructed, wherein the data set Ck includes: the top floor area difference t1 between the topographic map and the cadastral map, the second top floor area difference t2 between the topographic map and the cadastral map, and construct sets [t1] and [t2] for the top floor and second top floor differences of each house in an area;

[0015] S404: assign a weight to the data difference construction sequence Ck, reorder the set Ck, arrange the data according to the house location to obtain the sets [t1`] and [t2`], and calculate the adjacent ratios H1 and H2 of the top floor area difference t1 and the second top floor area difference t2. = , = , and are the k-th and k+1-th elements in the sequence [t1`], and For the k-th and k+1-th elements in the sequence [t2`], the values ​​of the adjacent ratios H1 and H2 are determined. If the values ​​of the adjacent ratios H1 and H2 are positive and greater than or equal to 1, the adjacent ratios are defined as forward weights. If the values ​​of the adjacent ratios H1 and H2 are positive and less than 1, the adjacent ratios are defined as reverse weights. The direction is determined by the forward weight and the reverse weight. In the data difference construction sequence Ck, the forward weight direction is from the k-th element to the k+1-th element, and the reverse weight direction is from the k+1-th element to the k-th element. The direction of the adjacent ratios H1 and H2 is assigned based on the direction of the edge line in step S401. Then the edge line direction of the kth element in the sequence Ck constructed by data difference is the adjacent ratio direction, and the weight vectors of the adjacent ratios H1 and H2 are calculated. , ,in, and is the weight vector of adjacent ratios H1 and H2, , are the vector values ​​of adjacent ratios after assigning directions to adjacent ratios H1 and H2 respectively, and m is the total number of elements of the sequence Ck constructed by data differences.

[0016] Furthermore, a picture comparison prediction model is established to optimize the overlay comparison analysis of house topographic maps and cadastral maps;

[0017] The image contrast information is optimized through the forward weight direction and the reverse weight direction, and the contrast image information is input into the neural network. The network output result is determined by the forward output and the reverse output. The calculation formula is as follows:

[0018] ;

[0019] ;

[0020] ;

[0021] ;

[0022] Among them, x represents the input value, and the input value is the sequence Ck constructed by data difference. are the forward weight and reverse weight directions respectively, is the positive output value of the network, , , are the input layer weight, hidden layer weight and bias vector of the forward output respectively. The input layer weight matrix is ​​constructed by the weight ratios H1 and H2 of the antenna area difference t1 and the secondary antenna area difference t2. is the reverse output value of the network, , , are the input layer weight, hidden layer weight and bias vector of the reverse output respectively. The input layer weight is the result obtained by comparing the top layer when judging the forward output and the reverse output. L1 is the input layer weight, L2 is the hidden layer weight, and the bias vector is and If the result is obtained by comparing the second top layer, L2 is the input layer weight, L1 is the hidden layer weight, and the bias vector is and of and;

[0023] is the final output value, and are the output layer weight matrix and bias vector, respectively, and the symbols For splicing operation;

[0024] The final output value Put it into the neural network model for training. After the training is completed, the input data is quickly output as an indicator X. The indicator X is used to judge whether the comparison data meets the qualified standard here. Each indicator X needs to be trained through a separate neural network model to obtain the mapping relationship between the input data and the indicator;

[0025] Among them, in the neural network model, the number of neurons in the network input layer is equal to the number of input data types, the number of neurons in the output layer is equal to the number of output data types, and the number of intermediate layers and the number of neurons in each layer are freely set. The setting of the intermediate layer will directly affect the accuracy of the approximate data of the neural network. Whether the number of intermediate layers or neurons needs to be increased or decreased is determined based on the training results. The calculation method for determining the number of intermediate layers and the number of neurons is:

[0026] The middle layers in the neural network model are added layer by layer, and the output index X is named according to the number of middle layers. When the number of middle layers is 1, the output index , and so on, and according to the indicator X, construct a nonlinear regression function, f (X) = , a is a constant value, calculate the variance D of the function's independent variable, D= , where L is the number of intermediate layers at output, is the i-th index X of the output, is the input data corresponding to the output index X, when =0, A takes the minimum value, where () is to find the partial derivative of the function, A is a parameter to determine the number of intermediate layers, and the value of A is compared with the variance D. If D>A, the output index X is input into the neural network model, and the number of intermediate layers is increased by 1. If D≤A, the output index X is output;

[0027] Update the weight coefficients of neurons by gradient descent:

[0028] (l) = - ;

[0029] in, () is the partial derivative of the function, P and Q are the system output error and neuron weight increment, both are constant values. (l) is the neuron weight coefficient, (l) is the updated neuron weight coefficient, Represents the neuron learning rate, which is (l) Determine the number of neurons to be set in the middle layer and combine the output values ​​according to the superposition analysis algorithm The algorithm is deeply learned to output top-level comparative analysis data and second-level comparative analysis data to ensure the quality of the comparative data.

[0030] Further, the house topographic map sky surface is fused with the house topographic map sub-sky surface, and the cadastral map sky surface is fused with the cadastral map sub-sky surface, and the fused sky surface comparison data is output;

[0031] Obtain the quadrant coordinates of the natural building of the integrated house and land corresponding to the house topographic sweet surface map, compare the quadrant coordinates of the natural building of the integrated house and land by fusing the house topographic map sky surface and the house topographic map sub-sky surface, and output the quadrant coordinate comparison data;

[0032] The auxiliary line coordinates of the residential area are obtained through the boundary line of the house, and the auxiliary line coordinate comparison data is output by comparing with the house topographic map.

[0033] By comparing the quadrant coordinates of the natural building integrated with the land with the coordinates of the auxiliary lines of the residential area, the coordinate comparison data of the natural building and the auxiliary lines of the land are output;

[0034] The fused celestial surface comparison data, quadrant coordinate comparison data, ground auxiliary line coordinate comparison data, natural building and ground auxiliary line coordinate comparison data are combined with the top-level comparison analysis data and the second-top-level comparison analysis data output by the overlay analysis algorithm to construct a database, and the comparison analysis report is output through the database.

[0035] Furthermore, the method is based on a system, and the system includes: a processor and a memory. When the processor executes the computer program, it can implement the steps in a method for improving the quality of integrated real estate data based on multi-source data as described in any one of the above methods.

[0036] The beneficial effects of the present invention are as follows: by overlaying and analyzing the topographic map and the cadastral map, the data in the original database of the integrated real estate and land is updated, so that the data quality of the original database is higher; and for the cadastral information and plot information in the integrated real estate and land, a more accurate comparative analysis report is output by comparing the areas of the topographic map and the cadastral map, making subsequent planning work more convenient. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The above and other features of the present invention will become more obvious by describing in detail the embodiments shown in the accompanying drawings. The same reference numerals in the accompanying drawings of the present invention represent the same or similar elements. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other accompanying drawings can be obtained based on these accompanying drawings without creative work. In the accompanying drawings:

[0038] Figure 1 Shown is an overall flow chart of a method for improving the quality of integrated real estate and land data based on multi-source data. DETAILED DESCRIPTION

[0039] The following will be combined with the embodiments and drawings to clearly and completely describe the concept, specific structure and technical effects of the present invention, so as to fully understand the purpose, scheme and effect of the present invention. It should be noted that the embodiments and features in the embodiments of this application can be combined with each other without conflict.

[0040] according to Figure 1 As shown, a method for improving the quality of real estate data based on multi-source data includes the following steps:

[0041] S100: Acquire a real-life picture of the house using a shooting device;

[0042] S200: Reading existing data in the database and constructing a three-dimensional model of the house according to the topographic map and the real-life picture of the house in the existing data;

[0043] S300: Acquire the real estate data, and acquire the cadastral map, the real estate natural building and the real estate plot map from the real estate data;

[0044] S400: Building a database by combining the three-dimensional model of the house with the real estate integrated data for comparative analysis, and outputting a comparative analysis report.

[0045] Furthermore, in step S100 and step S200, a real-life picture of the house is taken by a drone, the real-life picture of the house including: a roof picture and a side picture of the house, a three-dimensional model of the house is constructed by combining the house topographic map in a database with oblique photography, the roof picture of the house taken by the drone is mapped onto the top of the three-dimensional model of the house, and the roof picture of the house is combined with the topographic map taken by the drone to divide the roof of the house into a topographic top layer and a secondary top layer, and the elevations of various parts of the roof in the topographic top layer and the secondary top layer are determined, the building boundary line is determined according to the roof picture of the house, and a topographic roof picture of the house is obtained by comparing and analyzing the three-dimensional model of the house with the house topographic map, the topographic roof picture of the house being divided into a topographic top layer and a secondary top layer.

[0046] Furthermore, in step S300, a cadastral map, a natural building of integrated real estate and land, and a plot map of integrated real estate and land are obtained through a database, wherein the cadastral map indicates the ownership of the land in the topographic map, the natural building of integrated real estate and land indicates the exact location of the house in the integrated real estate and land system, and the plot map of integrated real estate and land reflects the basic situation of the plot, which includes: the plot ownership boundary, the location of boundary points, the location and nature of the buildings in the plot, and the baseline of the natural building of integrated real estate and land is determined according to the natural building of integrated real estate and land, and the partition ownership of the natural building is determined according to the baseline of the natural building.

[0047] Further, in step S400, the cadastral map part in the cadastral map is divided into a cadastral top layer and a cadastral second top layer, and the cadastral top layer and the cadastral second top layer are spatially superimposed and compared with the house topography top layer and the house topography second top layer, respectively. The specific method of superimposed and compared analysis includes the following steps:

[0048] S401: compare the top layer of the house terrain with the edge of the top layer of the cadastral map, and correct the edge of the top layer of the cadastral map by giving a higher weight to the top layer of the house terrain, so that the area of ​​the top layer of the cadastral map is recalculated, and the number of pixel points covered by the edge line is compared with the direction of the edge line through the edge line comparison to determine the correct direction of the edge line, and unify the cadastral information of the top layer sky map based on the correct direction of the edge line;

[0049] S402: Obtain the topographic sub-top floor of the house and the cadastral sub-top floor, determine the building range line of the edge part in the topographic sub-top floor of the house, re-determine the outer edge baseline of the cadastral sub-top floor according to the building range line, and determine the outer edge baseline according to the cadastral information of the top floor determined in step S401, determine the cadastral map top floor according to the outer edge baseline and the outer edge baseline, and perform weight matching on the area of ​​the cadastral sub-top floor determined according to the outer edge baseline and the inner edge baseline and the original cadastral sub-top floor area;

[0050] S403: The step of weighting the cadastral map area is as follows: record the area of ​​the cadastral second top floor determined by the external edge baseline and the internal edge baseline as CD1, define the original cadastral second top floor area as CD, the original cadastral second top floor area is CD, obtain the weight ratio of the original cadastral second top floor area as CD by comparing the comparison ratio of the topographic top floor and the cadastral top floor, and obtain the top floor area data difference in the building. The sequence Ck is constructed, wherein the data set Ck includes: the top floor area difference t1 between the topographic map and the cadastral map, the second top floor area difference t2 between the topographic map and the cadastral map, and construct sets [t1] and [t2] for the top floor and second top floor differences of each house in an area;

[0051] S404: assign a weight to the data difference construction sequence Ck, reorder the set Ck, arrange the data according to the house location to obtain the sets [t1`] and [t2`], and calculate the adjacent ratios H1 and H2 of the top floor area difference t1 and the second top floor area difference t2. = , = , and are the k-th and k+1-th elements in the sequence [t1`], and For the k-th and k+1-th elements in the sequence [t2`], the values ​​of the adjacent ratios H1 and H2 are determined. If the values ​​of the adjacent ratios H1 and H2 are positive and greater than or equal to 1, the adjacent ratios are defined as forward weights. If the values ​​of the adjacent ratios H1 and H2 are positive and less than 1, the adjacent ratios are defined as reverse weights. The direction is determined by the forward weight and the reverse weight. In the data difference construction sequence Ck, the forward weight direction is from the k-th element to the k+1-th element, and the reverse weight direction is from the k+1-th element to the k-th element. The direction of the adjacent ratios H1 and H2 is assigned based on the direction of the edge line in step S401. Then the edge line direction of the kth element in the sequence Ck constructed by data difference is the adjacent ratio direction, and the weight vectors of the adjacent ratios H1 and H2 are calculated. , ,in, and is the weight vector of adjacent ratios H1 and H2, , are the vector values ​​of adjacent ratios after assigning directions to adjacent ratios H1 and H2 respectively, and m is the total number of elements of the sequence Ck constructed by data differences.

[0052] The beneficial effect of determining the weight ratio in the step is: counting the differences and changes in the housing statistics in the area, and judging whether there are errors in the overall area of ​​the area based on the differences, and facilitating the subsequent planning based on the direction and size of the differences. The positive weight direction and the reverse weight direction are judged to be the influence and changes on the real estate data after the cadastral area is adjusted in a refined manner, and the range and direction of the changes need to be counted. Therefore, further comparison of the change amplitude and direction of adjacent houses is performed here, so that subsequent work can have clear data comparison.

[0053] Furthermore, a picture comparison prediction model is established to optimize the overlay comparison analysis of house topographic maps and cadastral maps;

[0054] The image contrast information is optimized through the forward weight direction and the reverse weight direction, and the contrast image information is input into the neural network. The network output result is determined by the forward output and the reverse output. The calculation formula is as follows:

[0055] ;

[0056] ;

[0057] ;

[0058] ;

[0059] Among them, x represents the input value, and the input value is the sequence Ck constructed by data difference. are the forward weight and reverse weight directions respectively, is the positive output value of the network, , , are the input layer weight, hidden layer weight and bias vector of the forward output respectively. The input layer weight matrix is ​​constructed by the weight ratios H1 and H2 of the antenna area difference t1 and the secondary antenna area difference t2. is the reverse output value of the network, , , are the input layer weight, hidden layer weight and bias vector of the reverse output respectively. The input layer weight is the result obtained by comparing the top layer when judging the forward output and the reverse output. L1 is the input layer weight, L2 is the hidden layer weight, and the bias vector is and If the result is obtained by comparing the second top layer, L2 is the input layer weight, L1 is the hidden layer weight, and the bias vector is and of and;

[0060] is the final output value, and are the output layer weight matrix and bias vector, respectively, and the symbols For splicing operation;

[0061] The final output value Put it into the neural network model for training. After the training is completed, the input data is quickly output as an indicator X. The indicator X is used to judge whether the comparison data meets the qualified standard here. Each indicator X needs to be trained through a separate neural network model to obtain the mapping relationship between the input data and the indicator;

[0062] The beneficial effect of the picture comparison prediction model in the step is that the neural network inputs the comparison image information through two-way comparison, and the accuracy of the comparison analysis is improved by performing a secondary analysis after a primary analysis, and the direction and size of the area difference can be recorded to reduce the computing power for subsequent comparisons. At the same time, the output value can be used as the output indicator X after training to perform deep learning on the comparison analysis, which greatly improves work efficiency.

[0063] Among them, in the neural network model, the number of neurons in the network input layer is equal to the number of input data types, the number of neurons in the output layer is equal to the number of output data types, and the number of intermediate layers and the number of neurons in each layer are freely set. The setting of the intermediate layer will directly affect the accuracy of the approximate data of the neural network. Whether the number of intermediate layers or neurons needs to be increased or decreased is determined based on the training results. The calculation method for determining the number of intermediate layers and the number of neurons is:

[0064] The middle layers in the neural network model are added layer by layer, and the output index X is named according to the number of middle layers. When the number of middle layers is 1, the output index , and so on, and according to the indicator X, construct a nonlinear regression function, f (X) = , a is a constant value, calculate the variance D of the function's independent variable, D= , where L is the number of intermediate layers at output, is the i-th index X of the output, is the input data corresponding to the output index X, when =0, A takes the minimum value, where () is to find the partial derivative of the function, A is a parameter to determine the number of intermediate layers, and the value of A is compared with the variance D. If D>A, the output index X is input into the neural network model, and the number of intermediate layers is increased by 1. If D≤A, the output index X is output;

[0065] Update the weight coefficients of neurons by gradient descent:

[0066] (l) = - ;

[0067] in, () is the partial derivative of the function, P and Q are the system output error and neuron weight increment, both are constant values. (l) is the neuron weight coefficient, (l) is the updated neuron weight coefficient, Represents the neuron learning rate, which is (l) Determine the number of neurons to be set in the middle layer and combine the output values ​​according to the superposition analysis algorithm The algorithm is deeply learned to output top-level comparative analysis data and second-level comparative analysis data to ensure the quality of the comparative data.

[0068] Further, the house topographic map sky surface is fused with the house topographic map sub-sky surface, and the cadastral map sky surface is fused with the cadastral map sub-sky surface, and the fused sky surface comparison data is output;

[0069] Obtain the quadrant coordinates of the natural building of the integrated house and land corresponding to the house topographic sweet surface map, compare the quadrant coordinates of the natural building of the integrated house and land by fusing the house topographic map sky surface and the house topographic map sub-sky surface, and output the quadrant coordinate comparison data;

[0070] The auxiliary line coordinates of the residential area are obtained through the boundary line of the house, and the auxiliary line coordinate comparison data is output by comparing with the house topographic map.

[0071] By comparing the quadrant coordinates of the natural building integrated with the land with the coordinates of the auxiliary lines of the residential area, the coordinate comparison data of the natural building and the auxiliary lines of the land are output;

[0072] The fused celestial surface comparison data, quadrant coordinate comparison data, ground auxiliary line coordinate comparison data, natural building and ground auxiliary line coordinate comparison data are combined with the top-level comparison analysis data and the second-top-level comparison analysis data output by the overlay analysis algorithm to construct a database, and the comparison analysis report is output through the database.

[0073] Furthermore, the method is based on a system, and the system includes: a processor and a memory. When the processor executes the computer program, it can implement the steps in a method for improving the quality of integrated real estate data based on multi-source data as described in any one of the above methods.

[0074] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete component gate circuits or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the system, and various sub-areas of the entire system are connected using various interfaces and lines.

[0075] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the plant control system based on artificial neural network by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0076] Although the description of the present invention has been quite detailed and has been described in particular with respect to several described embodiments, it is not intended to be limited to any of these details or embodiments or any particular embodiment, so as to effectively cover the intended scope of the present invention. In addition, the present invention is described above with the embodiments foreseeable by the inventors, and its purpose is to provide a useful description, and those non-substantial changes to the present invention that are not currently foreseen may still represent equivalent changes of the present invention.

Claims

1. A method for improving the quality of real estate data based on multi-source data, characterized in that: The method comprises the following steps: S100: Acquire a real-life picture of the house using a shooting device; S200: Reading existing data in the database and constructing a three-dimensional model of the house according to the topographic map and the real-life picture of the house in the existing data; S300: Acquire the real estate data, and acquire the cadastral map, the real estate natural building and the real estate plot map from the real estate data; S400: Building a database by combining the three-dimensional model of the house with the real estate integrated data for comparative analysis, and outputting a comparative analysis report; In step S400, the cadastral map part in the cadastral map is divided into a cadastral top layer and a cadastral second top layer, and the cadastral top layer and the cadastral second top layer are spatially superimposed and compared with the house topography top layer and the house topography second top layer respectively. The specific method of spatial superimposition and comparison analysis includes the following steps: S401: Compare the top layer of the house terrain with the edge of the top layer of the cadastral map, and correct the edge of the top layer of the cadastral map by giving a higher weight to the top layer of the house terrain, so that the area of ​​the top layer of the cadastral map is recalculated, and the number of pixel points covered by the edge line is compared with the direction of the edge line through edge line comparison to determine the correct direction of the edge line, and unify the cadastral information of the top layer sky map based on the correct direction of the edge line; S402: Obtain the topographic sub-top floor of the house and the cadastral sub-top floor, determine the building range line of the edge part in the topographic sub-top floor of the house, re-determine the outer edge baseline of the cadastral sub-top floor according to the building range line, and determine the outer edge baseline according to the cadastral information of the top floor determined in step S401, determine the cadastral map top floor according to the outer edge baseline and the outer edge baseline, and perform weight matching on the area of ​​the cadastral sub-top floor determined according to the outer edge baseline and the inner edge baseline and the original cadastral sub-top floor area; S403: The step of weighting the area of ​​the cadastral second top floor determined by the external edge baseline and the internal edge baseline and the original cadastral second top floor is as follows: record the area of ​​the cadastral second top floor determined by the external edge baseline and the internal edge baseline as CD1, define the original cadastral second top floor area as CD, obtain the weight ratio of the original cadastral second top floor area CD by comparing the comparison ratio of the topographic top floor and the cadastral top floor, and obtain a sequence Ck for the top floor area data difference in the building, wherein the sequence Ck includes: the top floor area difference t1 between the topographic map and the cadastral map, the second top floor area difference t2 between the topographic map and the cadastral map, and construct sets [t1] and [t2] for the top floor and second top floor differences of each house in an area; S404: assign weights to the sequence Ck constructed by data differences, reorder the sequence Ck, arrange the data according to the house locations to obtain sets [t1`], [t2`], and calculate the adjacent ratios H1, H2 of the top floor area difference t1 and the second top floor area difference t2. = , = , is the kth element in the sequence [t1`], is the k+1th element in the sequence [t1`], is the kth element in the sequence [t2`], For the k+1th element in the sequence [t2`], the values ​​of the adjacent ratios H1 and H2 are determined. If the values ​​of the adjacent ratios H1 and H2 are positive and greater than or equal to 1, the adjacent ratios are defined as positive weights. If the values ​​of the adjacent ratios H1 and H2 are positive and less than 1, the adjacent ratios are defined as reverse weights. The direction is determined by the forward weight and the reverse weight. In the data difference construction sequence Ck, the forward weight direction points from the kth element to the k+1th element, and the reverse weight direction points from the k+1th element to the kth element. The direction of the adjacent ratios H1 and H2 is given based on the direction of the edge line in step S401. Then the edge line direction of the kth element in the sequence Ck constructed by data difference is the adjacent ratio direction, and the weight vectors of the adjacent ratios H1 and H2 are calculated. , ,in, and is the weight vector of adjacent ratios H1 and H2, , are the vector values ​​of the adjacent ratios after the directions of the adjacent ratios H1 and H2 are assigned respectively, and m is the total number of elements of the sequence Ck constructed by the data difference; Establish a picture comparison prediction model to optimize the overlay comparison analysis of house topographic maps and cadastral maps; The image contrast information is optimized by the forward weight direction and the reverse weight direction, and the contrast image information is input into the neural network. The network output result is determined by the forward output and the reverse output. The calculation formula is as follows: ； ； ； ； Among them, x represents the input value, and the input value is the sequence Ck constructed by data difference. are the forward weight direction and the reverse weight direction respectively, is the positive output value of the network, , , are the input layer weights, hidden layer weights and bias vectors of the forward output respectively. The input layer weight matrix is ​​constructed by the weight ratios H1 and H2 of the antenna area difference t1 and the secondary antenna area difference t2. is the reverse output value of the network, , , are the input layer weight, hidden layer weight and bias vector of the reverse output respectively. The input layer weight is obtained by comparing the forward output and the reverse output. If the result is obtained by the top layer comparison, L1 is the input layer weight, L2 is the hidden layer weight, and the bias vector is and If the result is obtained by comparing the second top layer, L2 is the input layer weight, L1 is the hidden layer weight, and the bias vector is and of and; is the final output value, and are the output layer weight matrix and bias vector, respectively, and the symbols For splicing operation; The final output value The neural network model is put into training, and after the training is completed, the input data is quickly outputted as an indicator X, where the indicator X is used to judge whether the comparison data meets the qualified standard. Each indicator X needs to be trained through a separate neural network model to obtain the mapping relationship between the input data and the indicator; Among them, in the neural network model, the number of neurons in the network input layer is equal to the number of input data types, the number of neurons in the output layer is equal to the number of output data types, and the number of intermediate layers and the number of neurons in each layer are freely set. The setting of the intermediate layer will directly affect the accuracy of the approximate data of the neural network. According to the training results, it is judged whether the number of intermediate layers or neurons needs to be increased or decreased. The calculation method for judging the number of intermediate layers and neurons is: The middle layers in the neural network model are added layer by layer, and the output index X is named according to the number of middle layers. When the number of middle layers is 1, the output index , and so on, and according to the indicator X, construct a nonlinear regression function, f (X) = , a is a constant value, calculate the variance D of the function's independent variable, D = , where L is the number of intermediate layers at output, is the i-th index X of the output, is the input data corresponding to the output index X, when =0, A takes the minimum value, where () is to find the partial derivative of the function, A is a parameter to determine the number of intermediate layers, and the value of A is compared with the variance D. If D>A, the output index X is input into the neural network model, and the number of intermediate layers is increased by 1. If D≤A, the output index X is output; Update the weight coefficients of neurons by gradient descent: （l）=- ； in, () is the partial derivative of the function, P and Q are the system output error and neuron weight increment, both are constant values. (l) is the neuron weight coefficient, (l) is the updated neuron weight coefficient, Represents the neuron learning rate, which is (l) Determine the number of neurons to be set in the middle layer and combine the output values ​​according to the superposition analysis algorithm The algorithm is deeply learned to output top-level comparative analysis data and second-level comparative analysis data to ensure the quality of the comparative data.

2. According to the method for improving the quality of real estate data based on multi-source data in claim 1, it is characterized in that: In step S100 and step S200, a real-life picture of the house is taken by a drone, the real-life picture of the house including: a roof picture and a side picture of the house, a three-dimensional model of the house is constructed by combining a house topographic map in a database with oblique photography, the roof picture of the house taken by the drone is mapped onto the top of the three-dimensional model of the house, and the roof picture of the house is combined with the topographic map taken by the drone to divide the roof of the house into a topographic top layer and a second topographic top layer, and the elevations of various parts of the roof in the topographic top layer and the second top layer are determined, the building boundary line is determined according to the roof picture of the house, and a topographic roof picture of the house is obtained by comparing and analyzing the three-dimensional model of the house with the house topographic map, the topographic roof picture of the house is divided into a topographic top layer and a second topographic top layer.

3. According to the method for improving the quality of real estate data based on multi-source data in claim 1, it is characterized in that: In step S300, the cadastral map, the integrated natural building and the integrated real estate map are obtained through the database, the cadastral map indicates the ownership of the land in the topographic map, the integrated natural building indicates the accurate position of the house in the integrated real estate system, the integrated real estate map reflects the basic situation of the land, and the basic situation of the land includes: the land ownership boundary, the location of the boundary points, the location and nature of the buildings in the land, and the integrated natural building baseline of the natural building is determined according to the integrated natural building, and the partition ownership of the natural building is determined according to the natural building baseline.

4. According to the method for improving the quality of real estate data based on multi-source data in claim 1, it is characterized in that: The house topographic map sky surface is merged with the house topographic map sub-sky surface, and the cadastral map sky surface is merged with the cadastral map sub-sky surface, and the fused sky surface comparison data is output; Obtain the quadrant coordinates of the natural building of the integrated house and land corresponding to the house topographic sweet surface map, compare the quadrant coordinates of the natural building of the integrated house and land by fusing the house topographic map sky surface and the house topographic map sub-sky surface, and output the quadrant coordinate comparison data; The auxiliary line coordinates of the residential area are obtained through the boundary line of the house, and the auxiliary line coordinate comparison data is output by comparing with the house topographic map. By comparing the quadrant coordinates of the natural building integrated with the land with the coordinates of the auxiliary lines of the residential area, the coordinate comparison data of the natural building and the auxiliary lines of the land are output; The fused celestial surface comparison data, quadrant coordinate comparison data, ground auxiliary line coordinate comparison data, natural building and ground auxiliary line coordinate comparison data are combined with the top-level comparison analysis data and the second-top-level comparison analysis data output by the overlay analysis algorithm to construct a database, and the comparison analysis report is output through the database.

5. According to the method for improving the quality of real estate data based on multi-source data in claim 1, it is characterized in that: The method is based on a system, which includes: a processor and a memory. When the processor executes the computer program, it can implement the steps in the method for improving the quality of integrated real estate data based on multi-source data as described in any one of claims 1 to 4.

Images