CN110264381B - A method for estimating housing occupancy rate - Google Patents

Abstract

The invention belongs to the technical field of data processing, and particularly relates to a house occupancy rate estimation method. The invention creatively utilizes the brightness of the lamplight to calculate the residence ratio of the house, comprehensively considers the two-dimensional and three-dimensional structural information of the residential building, and provides an estimation method of the residence ratio of the house. The invention determines the residential building proportion of each grid unit based on residential building data, and carries out residential building partition; the mixed pixel idea is adopted to remove the brightness of the non-residential lights, so that the error caused by removing the non-residential lights by adopting a unique value can be effectively reduced; the house residence ratio is calculated in a partitioning way by adopting the partition thought, so that the simulation precision of the existing method can be effectively improved. In addition, the average relative error (6.33%) of the method is obviously lower than that of the existing estimation method (33.83%), and the method can be used for rapidly, conveniently and accurately acquiring the residence ratio of the house. In addition, the method has low requirement on time limit, and can be applied to data in different periods.

Description

A method for estimating housing occupancy rate

Technical field

The invention belongs to the field of data processing technology, and specifically relates to a method for estimating house occupancy rate.

Background technique

Housing occupancy rate is an important indicator reflecting the real occupancy status of a house. The acceleration of urbanization has greatly promoted the development of the real estate industry. Accurately grasping the spatial distribution of urban housing occupancy rates is of great significance to the coordinated and orderly development of the real estate industry and the national economy. With the development of remote sensing and GIS technology, the application of remote sensing and GIS technology to monitor housing occupancy has become an important means to quickly, conveniently and accurately obtain the true occupancy status of houses.

Traditional estimates of housing occupancy rates are mainly obtained through statistical data, real estate statistical information and questionnaire surveys. For example, developed countries use a housing vacancy indicator system to reflect housing usage and have authoritative survey agencies. For example, the U.S. Census Bureau has three types of housing surveys: the American Social Survey (ACS) and the Current Population Survey. The Survey/Housing Vacancy Survey (CPS/HVS) and the American Housing Survey (AHS) mainly conduct surveys and analyzes of housing stock. However, this method is time-consuming and labor-intensive, cannot obtain the internal spatial distribution information of housing occupancy rates, and different countries The statistical caliber and statistical scale are also different. Real estate vacancy rate surveys are generally statistics compiled by real estate companies on the number of unsold residential units, the number of unoccupied residential units, and the number of occupied and unoccupied residential units. They are generally carried out on a small scale and cannot be applied on a large scale.

Moreover, in addition to the disadvantages of differences in statistical caliber and scale, time-consuming and labor-intensive research, the above methods often cannot be applied on a large scale, and cannot obtain internal spatial distribution information of housing occupancy rates. Although obtaining the housing vacancy rate based on remote sensing and GIS methods provides a solution for simulating the spatial distribution of housing occupancy rates, this type of method has shortcomings such as the inability to accurately identify the proportions of residential buildings and the failure to consider the three-dimensional structure information of residential buildings, which affects the accuracy of the simulation. certain influence.

Therefore, there is an urgent need to study an estimation method of housing occupancy rate that can improve the above technical problems.

Contents of the invention

In order to improve the above technical problems, the present invention provides a method for estimating housing occupancy rate. The estimation method includes the following steps:

1) Night lighting data preprocessing;

2) The ratio of the area of residential buildings to non-residential areas is determined;

3) Residential area zoning;

4) Extraction of light brightness in residential areas;

5) Determine the lighting brightness of full residential areas in different zones;

6) Calculation of housing occupancy rate.

In step 1), the preprocessing is to remove background noise, which includes short-term light data such as forest fires, auroras, and volcanoes, as well as snow on mountain tops, dry beds, etc. The preprocessing uses an eight-neighbor algorithm to smooth the nighttime light data. To facilitate calculation, the nighttime light data is resampled to 450m×450m.

In step 2), since the construction land consists of residential buildings and non-residential areas, the proportion of non-residential areas can be determined.

The construction land can be extracted from any of land use data, land cover, impervious surface and other data.

For example, based on land use data, urban construction land is extracted, and a 450m × 450m fishing net is used as a statistical unit to calculate the proportion of construction land area in each fishing net. Taking the 450m × 450m fishing net as the statistical unit, the non-residential area ratio is the construction land area ratio within the fishing net unit minus the residential building area ratio within the fishing net unit;

The non-residential area ratio is determined as shown in the following formula (1):

(1)

In the formula, is the area ratio of non-residential areas,/> is the construction land area within the fishing net unit,/> is the unit area of the fishing net,/> It is the proportion of residential building area in the fishing net unit.

Specifically, based on the residential building data, taking the 450m × 450m fishing net as the statistical unit, the proportion of residential building area and the proportion of residential building area for different floors in each fishing net unit are calculated. The calculation formulas are as shown in Equations (2) and (3):

In the formula, For the fishing net unit/> Number of floors Ratio of residential building area,/> is the number of floors,/> ,/> Respectively, they are No./> in the fishing net unit. Number of floors, building area and fishing net unit area;/> It is the highest number of floors in the fishing net unit.

In step 3), the residential area zoning is specifically: based on the residential building area ratio of different floors of each fishing net unit, the residential area is divided according to the classification standards of low-rise, middle-rise, mid-rise and high-rise, and the low-rise is 1-n1 layer, the middle layer is the n2-n3 layer, the middle and high layer is the n4-n5 layer, and the high layer is the n6-n layer; for example, the low layer is the 1-3 layer, and the middle layer is the 4-6 layer, The middle and high-rise buildings are 7-9 floors, and the high-rise buildings are >9 floors. The n1, n2, n3, n4, n5 and n6 are selected from integers greater than 1; the calculation formula of the residential area division is as follows: formula (4 ) and shown in equation (5):

In the formula, It is the highest proportion of residential floor area in the fishing net unit,/> It is a residential building category for fishing net units.

In step 4), the light brightness of the residential area is extracted. Since nighttime lights not only include residential building lights, but may also include stray light from roads, commercial areas, parks, etc., therefore, in order to obtain pure lighting for residential buildings in the fishing net unit, the non-residential area lights in the fishing net unit need to be eliminated. Using the idea of mixed pixels, the calculation formula of the light brightness in the residential area is:

In the formula, It is the average light brightness per unit area of non-residential area,/> is the brightness of the light in the residential area,/> is the original light brightness.

Further, the It can be calculated by the following method: select several (for example, 100-300, for example, 200) pixels that do not contain residential buildings based on high-resolution remote sensing images, and extract the light brightness value of the sample; then, calculate the unit non-residential area The light brightness corresponding to the area ratio is determined by averaging the light brightness per unit non-residential area area ratio. The calculation formula is as shown in Equation (7).

In the formula, Select sample size for non-residential areas,/> For the first/> The light brightness corresponding to each sample.

In step 5), determine the light brightness of the full residential area in different zones. The specific step 5) is: determining the lighting brightness of the full residential area in each partition.

Furthermore, based on step 3) residential area zoning results and step 4) residential area light brightness extraction results, statistics can be made on the light brightness frequency histograms of each zone. To facilitate calculation, the light values of each pixel can be rounded. The light brightness of the full residential area in different partitions can be selected based on the frequency histogram to select the light brightness with a cumulative frequency of 80% as the light brightness of the full residential area in the partition.

In step 6), the occupancy rate of the house can be determined based on the residential light ratio. The calculation formula of the housing occupancy rate is shown in Equation (8).

In the formula, For the first/> The housing occupancy rate of each fishing net unit and the lighting brightness of the residential area,/> It is the lighting brightness that fills the living area.

beneficial effects

The present invention creatively uses light brightness to calculate the house occupancy rate, and comprehensively considers the two-dimensional and three-dimensional structural information of residential buildings to provide a method for estimating the house occupancy rate. This invention determines the proportion of residential buildings in each grid unit based on residential building data, and partitions residential buildings; it uses the idea of mixed pixels to remove the brightness of non-residential lights, which can effectively reduce the error caused by using unique values to remove lights in non-residential areas; it adopts partitioning The idea is to calculate the housing occupancy rate by partition, which can effectively improve the simulation accuracy of existing methods. In addition, the average relative error of the method of the present invention (6.33%) is significantly lower than the existing estimation method (33.83%). The method of the present invention can quickly, conveniently and accurately obtain the house occupancy rate.

In addition, the method of the present invention does not have high time limit requirements and can be applied to data in different periods.

Description of drawings

Figure 1 is the overall technical flow chart of the housing occupancy rate estimation method.

Figure 2 is a processing flow chart of step 1) of the present invention.

Figure 3 is a processing flow chart of step 2) of this application.

Figure 4 is a processing flow chart of step 3) of this application.

Figure 5 is a processing flow chart of step 4) of this application.

Figure 6 is a processing flow chart of step 5) of this application.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments These are some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

A method for estimating housing occupancy rate, including the following steps:

Step 1: Night light data preprocessing. Since the NPP-VIIRS nighttime light data contains short-term light data such as forest fires, auroras, and volcanoes, as well as background noise from mountaintop snow and dry beds, the data needs to be preprocessed to eliminate the influence of noise. The processing flow is shown in Figure 2. The preprocessing scheme uses the eight-neighbor algorithm to smooth the nighttime light data. To facilitate calculation, the nighttime light data is resampled to 450m×450m.

Step 2: Determine the area ratio of residential buildings to non-residential areas. The processing flow is shown in Figure 3. The residential building patch area is the bottom area of each residential building element. Based on the residential building data, the 450m×450m fishing net is used as the statistical unit to calculate the proportion of residential building area and the proportion of residential building area of different floors in each fishing net unit. The calculation formulas are as shown in Equations (2) and (3). Based on the land use data, urban construction land was extracted, and the 450m×450m fishing net was used as the statistical unit to calculate the proportion of construction land area in each fishing net. Since construction land is composed of residential buildings and non-residential areas, the area ratio of non-residential areas can be determined, and the calculation formula is as shown in Equation (1).

In the formula, For the fishing net unit/> Number of floors Ratio of residential building area,/> is the number of floors,/> ,/> Respectively, they are No./> in the fishing net unit. Number of floors, building area and fishing net unit area;/> is the proportion of residential building area in the fishing net unit,/> is the highest number of floors in the fishing net unit;/> is the area ratio of non-residential areas,/> It is the construction land area within the fishing net unit.

Step 3: Residential area zoning. The processing flow is shown in Figure 4. Based on the residential building area ratio of different floors of each fishing net unit obtained in step 2, according to the low-rise (1-3 floors), middle-floor (4-6 floors), mid-high-rise (7-9 floors) and The classification standard for high-rise buildings (>9 floors) divides residential areas, and the calculation formulas are as shown in Equations (4) and (5):

In the formula, It is the highest proportion of residential floor area in the fishing net unit,/> It is a residential building category for fishing net units.

Step 4: Extraction of light brightness in residential areas. The processing flow is shown in Figure 5. Since nighttime lights not only include residential building lights, but may also include stray light from roads, commercial areas, parks, etc., therefore, in order to obtain pure lighting for residential buildings in the fishing net unit, it is necessary to analyze the non-linear lighting in the fishing net unit. Residential area lights are removed. Because the light brightness in urban areas and suburbs is different, we believe that when the proportion of construction land in a fishing net unit is 100%, the fishing net unit is located in the urban area. When it is lower than 100%, it is in the suburbs. The lighting brightness of residential areas is determined according to urban areas and suburbs. First, based on high-resolution remote sensing images, 200 pixels that do not contain residential buildings were randomly selected from urban and suburban areas as non-residential building area samples, and the light brightness values of the samples were extracted; then, the corresponding unit non-residential area area ratio was calculated The average light brightness of the unit non-residential area ratio is determined by averaging. The calculation formula is as shown in Equation (7). Finally, the mixed pixel idea is used to calculate the urban and suburban grid units according to Equation (6). Light brightness in inner living areas.

In the formula, It is the average light brightness per unit area of non-residential area,/> Select sample size for non-residential areas,/> For the first/> The light brightness corresponding to each sample;/> is the brightness of the light in the residential area,/> is the original light brightness.

Step 5: Determine the lighting brightness of full living areas in different zones. The processing flow is shown in Figure 6. In order to calculate the occupancy rate, it is first necessary to determine the light brightness of the full residential area in each zone. Based on the residential area zoning results in step 3 and the residential area light brightness extraction results in step 4, statistics are made on the light brightness frequency histograms of each zone. To facilitate calculation, the light values of each pixel are rounded. According to the frequency histogram, the light brightness with a cumulative frequency of 80% is selected as the light brightness of the full residential area in the partition.

Step 6: House occupancy rate calculation. Based on the light brightness of the full residential area determined by the partition in step 5, the housing occupancy rate of each fishing net unit is calculated respectively. The calculation formula is shown in Equation (8).

In the formula, and/> Respectively:/> The housing occupancy rate of each fishing net unit and the lighting brightness of the residential area,/> It is the lighting brightness that fills the living area.

The average relative error of the method of the present invention is 6.33%.

Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent substitutions for some of the technical features; and these Modifications or substitutions do not deviate from the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims (6)

1. A method for estimating house occupancy rate, characterized in that the estimation method includes a step of calculating house occupancy rate: the ratio of the light brightness of the residential area to the light brightness of the full residential area is the house occupancy rate; The estimation method includes the following steps: 1) Night lighting data preprocessing; 2) The ratio of the area of residential buildings to non-residential areas is determined; 3) Residential area zoning; Specifically: based on the proportion of the residential building area of different floors in each fishing net unit, the residential area is divided according to the classification standards of low-rise, middle-rise, mid-rise and high-rise. The low-rise is the 1-n1 floor, and the middle-rise is the n2-n3 floor. , the middle and high-rise buildings are n4-n5 floors, the high-rise buildings are n6-n floors, and the n1, n2, n3, n4, n5 and n6 are selected from integers greater than 1; the calculation formula of the residential area division is as follows: (4) and equation (5) are shown: In the formula, For the fishing net unit/> Number of floors Ratio of residential building area,/> is the number of floors,/> It is the highest proportion of residential floor area in the fishing net unit,/> It is a residential building category for fishing net units; 4) Extraction of light brightness in residential areas; Among them, the calculation formula of the light brightness in the residential area is: In the formula, It is the average light brightness per unit area of non-residential area,/> is the brightness of the lighting in the residential area, is the original light brightness,/> is the proportion of non-residential area; described The calculation formula is shown in Equation (7): In the formula, Select sample size for non-residential areas,/> For the first/> The light brightness corresponding to each sample; 5) Determine the lighting brightness of full residential areas in different zones; 6) Calculation of housing occupancy rate. 2. The method for estimating house occupancy rate according to claim 1, characterized in that in step 1), the preprocessing is to remove background noise, and the background noise includes short-term light data of forest fires, aurora, volcanoes and mountain tops. snow, dry beds; The preprocessing uses an eight-neighbor algorithm to smooth the nighttime light data. To facilitate calculation, the nighttime light data is resampled to 450m×450m. 3. The method for estimating the housing occupancy rate according to claim 1, characterized in that in step 2), the fishing net is used as a statistical unit, and the non-residential area area ratio is the construction land area ratio within the fishing net unit minus the fishing net unit. Proportion of internal residential building area; The non-residential area ratio is determined as shown in the following formula (1): (1) In the formula, is the construction land area within the fishing net unit,/> is the unit area of the fishing net,/> It is the proportion of residential building area in the fishing net unit. 4. The method for estimating the housing occupancy rate according to claim 3, characterized in that, in step 2), based on the residential building data, the fishing net is used as a statistical unit to calculate the proportion of residential building area and the number of residents in each fishing net unit with different floors. The building area ratio, the calculation formula is as shown in formula (2) and formula (3): In the formula, ,/> Respectively, they are No./> in the fishing net unit. Number of floors, building area and fishing net unit area;/> It is the highest number of floors in the fishing net unit. 5. The method for estimating the housing occupancy rate according to claim 1, characterized in that step 5) specifically includes: determining the light brightness of the full residential area in each partition; According to the frequency histogram, the light brightness with a cumulative frequency of 80% is selected as the light brightness of the fully occupied residential area in the partition. 6. The method for estimating the house occupancy rate according to claim 1, characterized in that in step 6), the calculation formula of the house occupancy rate is as shown in Equation (8): In the formula, For the first/> The housing occupancy rate of each fishing net unit and the lighting brightness of the residential area, It is the lighting brightness that fills the living area.