CN104851130B - A kind of three-dimensional generation method of satellite remote-sensing image - Google Patents

Abstract

A method for generating three-dimensional satellite remote sensing images, the steps of which are: (1) acquiring satellite remote sensing orthophotos; (2) using digital elevation model data to invert the slope angle and aspect of each pixel; In the image metafile, search for the solar elevation angle, azimuth angle and spatial resolution information of each pixel of the original satellite remote sensing image; (4) calculate the terrain generation factor of each pixel; (5) combine the terrain generation factor with the satellite The DN values of the corresponding pixels on the remote sensing orthophoto are multiplied to obtain the 3D satellite remote sensing image. The present invention uses the ground elevation information to calculate the real dark difference and shadow distribution of the satellite remote sensing image, inverts the terrain factor of each pixel and fits it with the satellite remote sensing image, so that the real surface height is integrated into the planar topography information Topographic information such as light and shade, so that the flat satellite remote sensing images can display three-dimensional effects at the same time, making it easier to identify landforms.

Description

A 3D Generation Method of Satellite Remote Sensing Image

technical field

The invention belongs to the field of remote sensing image processing and relates to a method for generating three-dimensional remote sensing images.

Background technique

Satellite remote sensing images have the ability to quickly obtain large-scale surface information, and can directly and truly reflect the comprehensive landscape characteristics of the surface, and have been widely used. However, satellite remote sensing images show two-dimensional landscapes, which cannot directly express the three-dimensional information of the surface undulations. How to endow planar images with the topography of the actual ups and downs of the surface, so that it has a three-dimensional simulation effect, is one of the difficulties in cartography.

At present, to obtain 3D images of satellite remote sensing images and make them have a three-dimensional effect, generally use the 3D display module similar to Google Earth in commercial software such as ERDAS and ARGIS. According to the digital elevation model (DEM) and the geographic coordinates of satellite remote sensing images Information matching, that is, based on the DEM displayed in three dimensions, the satellite remote sensing image is superimposed on the DEM according to the corresponding geographic coordinates for display. This method can only be demonstrated online in the system, and it is only a superimposed effect, not endow The satellite remote sensing image itself has three-dimensional topographic features, and it is impossible to generate satellite remote sensing pictures with three-dimensional topographic effects. The method of use is complicated and not very convenient.

Contents of the invention

The technical problem solved by the present invention is: to overcome the deficiencies of the prior art, to provide a three-dimensional generation method of satellite remote sensing images, to use the ground elevation information to calculate the real dark difference and shadow distribution of satellite remote sensing images during imaging, and to invert satellite remote sensing images. The terrain factor of each pixel of the image is then fitted with the satellite remote sensing image to endow the planar landform with real topographic information such as surface height, light and shade, and directly generate a planar satellite remote sensing image with a three-dimensional display effect, making the planar Satellite remote sensing images have a three-dimensional effect and can more intuitively reflect the real surface and landforms.

The technical solution of the present invention is: a kind of three-dimensional generation method of satellite remote sensing image, comprises the following steps:

(1) Obtain the original satellite remote sensing image, perform orthorectification on the original satellite remote sensing image, and obtain the satellite remote sensing orthophoto image;

(2) Use digital elevation model data to invert the slope angle β and slope aspect θ of each pixel in the original satellite remote sensing image;

(3) Find the solar elevation angle ω, azimuth α and spatial resolution value Δ of each pixel in the original satellite remote sensing image from the satellite remote sensing image meta-file;

(4) Using the results of step (3) and step (4), calculate the terrain generation factor S _BCGH of each pixel in the original satellite remote sensing image = Δ ² *(1+tanβ·cotω·cos(α-θ)) ;

(5) Multiply the terrain generation factor of each pixel with the DN value of the corresponding pixel on the satellite remote sensing orthophoto image in step (1) to obtain a three-dimensional satellite remote sensing image.

The calculation method of described slope angle β and slope aspect θ is:

Among them, P is the sampling interval of the ground elevation model, C ₁ , C ₂ , C ₃ , C ₄ , C ₅ , C ₆ , C ₇ , and C ₈ are the elevation values of the pixels adjacent to pixel C respectively, and the pixel The elevation value corresponding to the pixel directly above C is C ₂ , and then along the clockwise direction, the elevation values of the adjacent pixels to pixel C are C ₆ , C ₃ , C ₇ , C ₄ , C ₈ , C ₁ , _C5 .

Compared with the prior art, the present invention has the advantages that: the method of the present invention starts from the principle of satellite remote sensing image imaging and the perspective of three-dimensional visualization simulation of satellite remote sensing images, fully considers the influence of terrain on satellite imaging, and introduces terrain generation factors to the original satellite remote sensing Improving the image, simulating the performance of the difference in light and shade caused by the terrain on the satellite image, has a strong theoretical basis, and can generate a plane satellite image with a three-dimensional effect that has a significant three-dimensional effect and truly shows the topography. The operation is simple and convenient. It is flexible and convenient to use, and is easier to apply in actual business work. The image after 3D terrain simulation is significantly stronger than the original image, highlighting the spatial distribution of terrain ridgelines and valley lines, and can intuitively interpret landforms and terrains, with better visualization effects, while improving visual beauty and three-dimensionality , which has good application value in engineering route selection, geological environment investigation and other fields.

Description of drawings

Fig. 1 is a block flow diagram of the inventive method;

Fig. 2 is the distribution schematic diagram of adjacent pixel elevation value when calculating pixel slope angle and slope aspect for the present invention;

Fig. 3 is a schematic diagram of calculation of terrain generation factors in the present invention.

detailed description

As shown in Figure 1, it is a flow chart of the inventive method, and the main steps are as follows:

(1) Obtain the original satellite remote sensing image, perform orthorectification on the original satellite remote sensing image, and obtain the satellite remote sensing orthophoto.

For example, ERDAS software can be used to select the orthorectification module to open the original satellite remote sensing images and reference images, where the reference images can be topographic maps and other materials with precise positioning information. Select more than 20 control points with the same name on the original satellite remote sensing images and reference images, and perform orthorectification to obtain satellite remote sensing orthoimages with precise positioning.

(2) Using digital elevation model (DEM) data to invert slope angle and aspect.

Digital elevation model (Digital Elevation Model), referred to as DEM. It is a solid ground model that represents the ground elevation in the form of a set of ordered numerical arrays. Topographic features such as slope angle, slope aspect, and slope angle change rate can be calculated and inverted on the basis of DEM. Digital elevation model data is a data product generated by using DEM. Its expression is the same as that of satellite remote sensing images. Each pixel has geographic coordinate information, and the pixel value is the elevation value at this coordinate.

The slope angle reflects the inclination of the surface, which is defined as the angle between the normal direction of a point P on the surface and the vertical direction (ie, the zenith), while the slope aspect is the direction facing the slope, defined as the positive direction of the normal of P The angle between the projection on the plane and the direction of true north in a clockwise direction.

According to literature (Liu Xuejun et al. Calculation method of slope in any direction based on DEM [J]. Regional Research and Development, 2009 (04): 139-141.), for the point C( x, y), and the slope angle is the fastest descending value obtained along the opposite direction of gradient f(x, y) of C, and the fastest descending direction is the slope aspect.

The calculation formulas of slope angle β and slope aspect θ are:

θ=arctan(f _y /f _x )

Where f _x , f _y are the partial derivatives in the x and y directions, respectively.

In actual work, the slope aspect generally starts from the north direction and is measured in a clockwise direction. The slope aspect is expressed in a coordinate system where the x-axis is the north-south direction and the y-axis is the east-west direction:

θ＝270°+arctan(f _y /f _x )-90°*(f _x /|f _x |)

In DEM data, the calculation of f _x and f _y is generally in the local range (3*3 moving window, as shown in Figure 2), using numerical differentiation method or local surface fitting method, the calculation formula is as follows:

Among them, P is the sampling interval of digital elevation model data, which is recorded in the meta-file of digital elevation model data. Meta-files are generally stored in *.xml, *.txt and other formats, and the sampling interval of data can be directly read. C ₁ , C ₂ , C ₃ , C ₄ , C ₅ , C ₆ , C ₇ , and C ₈ are the elevation values of pixels adjacent to pixel C, respectively, and the specific distribution is shown in Figure 2.

From the above derivation, the calculation formulas of slope angle β and slope aspect θ among the present invention are as follows:

For cells at the edge of the image, missing values are replaced with a value of 0.

(3) Find the solar elevation angle ω and solar azimuth α when the scene image is imaged, and the spatial resolution Δ of the image from the satellite remote sensing image meta-file.

Each satellite remote sensing image contains a meta file, which is generally stored in *.xml, *.txt and other formats, and contains the basic information of the satellite remote sensing image, such as imaging time, latitude and longitude range, spatial resolution, and sun altitude angle and solar azimuth etc.

The sun altitude angle refers to the angle between the incident direction of sunlight and the ground plane. The solar azimuth refers to the angle between the projection of the sun's rays on the ground plane and the local meridian, which can be approximately regarded as the angle between the shadow of a straight line erected on the ground under the sun and the due south. Spatial resolution refers to the size of the ground range represented by a pixel, that is, the minimum distance between two adjacent ground objects that can be identified on satellite remote sensing images.

Due to the limited width of the satellite, the solar elevation angle and azimuth angle in one scene image are basically the same, so the solar elevation angle and azimuth angle of the center point of the image are simultaneously measured and recorded in the metadata of the satellite remote sensing image when imaging the satellite remote sensing image. As the solar elevation and azimuth values of each pixel in the scene image.

The spatial resolution of satellite remote sensing images, as an important parameter of satellite load, has been determined during the design of satellite sensors, and the spatial resolution of satellite remote sensing images acquired by the same sensor is the same.

(4) Calculate the terrain generation factor.

Due to the influence of terrain fluctuations, there will be differences in light and shade such as shady and sunny slopes on the surface, indicating that the direct solar radiation energy received by each pixel projected on the two-dimensional satellite remote sensing image is different. Conversely, the pixel The received direct solar radiation energy can also be an important factor in reflecting the topography. Therefore, the present invention uses the direct solar radiation energy received by the pixel to deduce the real light and shade difference of the ground surface during satellite imaging, to correct the flat satellite remote sensing image, endow the satellite remote sensing image with a real topographic environment, and thus highlight the three-dimensional display of the flat image The effect is to make the landform terrain more recognizable.

Since the direct solar radiation energy received by a satellite image pixel is proportional to its horizontal projected area in the direct sun direction, the present invention uses the horizontal projected area of each satellite image pixel in the direct sun direction as the terrain generation factor.

As shown in Figure 3, assuming that ABCD is the real surface of a certain pixel, which is located on a slope with slope angle β and slope aspect θ, its vertical projection BCEF is the pixel range on the satellite remote sensing image. The direct solar radiation energy is proportional to the horizontal projection BCGH of ABCD in the direct sun direction, and the area of BCGH is the terrain generation factor used in the present invention.

The area S _BCGH of BCGH is calculated as follows:

S _BCGH ＝S _BCEF *(1+tanβ·cotω·cos(α-θ))＝Δ ² *(1+tanβ·cotω·cos(α-θ)) where S _BCEF is the area of BCEF, β is ABCD , ω is the solar elevation angle, α is the solar azimuth, θ is the slope aspect of ABCD, S _BCEF is only related to the spatial resolution Δ of satellite remote sensing images, S _BCEF = Δ ² .

According to the above formulas, the terrain generation factor of each pixel in satellite remote sensing image imaging can be calculated.

(5) Multiply the terrain generation factor with the satellite remote sensing orthoimage to obtain the satellite remote sensing image after terrain generation.

Using ERDAS software, the terrain generation factor is multiplied by the original DN value of each pixel of the satellite remote sensing orthophoto image, and the image after terrain generation is obtained.

The content that is not described in detail in the description of the present invention belongs to the well-known technology of those skilled in the art.

Claims (1)

1. the three-dimensional generation method of a kind of satellite remote-sensing image, it is characterised in that comprise the following steps：

(1) original satellite remote sensing image is obtained, ortho-rectification is carried out to original satellite remote sensing image, obtains satellite remote sensing orthogonal projection Picture；

(2) the slope angle β and slope aspect θ of each pixel of Law of DEM Data inverting original satellite remote sensing image are utilized；

(3) from satellite remote-sensing image meta file, sun altitude ω, the side of each pixel of original satellite remote sensing image are searched Parallactic angle α and spatial resolution value, Δ；

(4) using step (2) and the result of step (3), the landform life of each pixel of original satellite remote sensing image is calculated Into factor S_BCGH=Δ²*(1+tanβ·cotω·cos(α-θ))；

(5) DN by the terrain generation factor of each pixel with the corresponding pixel on step (1) Satellite remote sensing orthography Value carries out multiplying, obtains three-dimensional satellite remote sensing image；

Described slope angle β and slope aspect θ computational methods are：

<mrow> <mi>&amp;beta;</mi> <mo>=</mo> <mi>arctan</mi> <msqrt> <mrow> <msup> <mrow> <mo>(</mo> <mfrac> <mrow> <mo>(</mo> <msub> <mi>C</mi> <mn>5</mn> </msub> <mo>+</mo> <msub> <mi>C</mi> <mn>2</mn> </msub> <mo>+</mo> <msub> <mi>C</mi> <mn>6</mn> </msub> <mo>)</mo> <mo>-</mo> <mo>(</mo> <msub> <mi>C</mi> <mn>8</mn> </msub> <mo>+</mo> <msub> <mi>C</mi> <mn>4</mn> </msub> <mo>+</mo> <msub> <mi>C</mi> <mn>7</mn> </msub> <mo>)</mo> </mrow> <mrow> <mn>6</mn> <mi>P</mi> </mrow> </mfrac> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>+</mo> <msup> <mrow> <mo>(</mo> <mfrac> <mrow> <mo>(</mo> <msub> <mi>C</mi> <mn>7</mn> </msub> <mo>+</mo> <msub> <mi>C</mi> <mn>3</mn> </msub> <mo>+</mo> <msub> <mi>C</mi> <mn>6</mn> </msub> <mo>)</mo> <mo>-</mo> <mo>(</mo> <msub> <mi>C</mi> <mn>8</mn> </msub> <mo>+</mo> <msub> <mi>C</mi> <mn>1</mn> </msub> <mo>+</mo> <msub> <mi>C</mi> <mn>5</mn> </msub> <mo>)</mo> </mrow> <mrow> <mn>6</mn> <mi>P</mi> </mrow> </mfrac> <mo>)</mo> </mrow> <mn>2</mn> </msup> </mrow> </msqrt> </mrow>

Wherein, P be ground elevation model sampling interval, C₁、C₂、C₃、C₄、C₅、C₆、C₇、C₈Respectively with pixel C adjacent picture elements Height value, height value corresponding to pixel is C directly over pixel C₂, then along clockwise direction, the height with pixel C adjacent picture elements Journey value is followed successively by C₆、C₃、C₇、C₄、C₈、C₁、C₅。