CN103186892A - Method and system for generating equal proportion live field map with aerial images - Google Patents

Abstract

The invention provides a method and a system for generating equal proportion live field map with aerial images. The method comprises the following steps of: building an aerial photography platform; photographing a vertical view of the site of accident; executing geometric correction on the vertical view; performing transformation of coordinates for the vertical view, and transforming the coordinate system of the vertical view into an equal proportion live coordinate system; and extracting and marking accident essential factor information. With the method and the system, a true and fine two-dimensional equal proportion live field map can be obtained through geometric correction and transformation of coordinates for the aerial images, and each essential factor of the site of accident is restored efficiently and accurately.

Description

Method and system for generating equal-scale real-scene scene map by using aerial images

technical field

The invention relates to the technical field of traffic accident scene information acquisition, in particular to a method and system for generating an equal-scale real-scene scene map by using aerial images.

Background technique

Accident site investigation and site map drawing are the first link in road traffic accident handling. The main tasks include the measurement and equal-scale drawing of various elements on the site, and taking pictures of the site, etc. In the shooting of the traffic accident scene, the camera is usually used for shooting, and each photo taken can only reflect the accident situation in one direction. To restore the accident scene, it is necessary to use the on-site measurement results to manually draw a schematic diagram on the coordinate paper. At present, the accident scene measurement is mainly carried out manually, so the manual drawing of the schematic diagram has the following limitations: (1) the efficiency is extremely low, and the measurement and drawing time is long, which is not conducive to the rapid recovery of the traffic accident scene; (2) the manual measurement accuracy is low, Moreover, manual drawing standards are different, and it is difficult to grasp the proportions of vehicles, roads, and traces; (3) some key parameters are easy to miss and cannot be made up; (4) it is impossible to perform secondary verification on the accident scene. In some disputed accidents, if there is doubt about some of the measured dimensions or there is a contradiction between the measured dimensions, the secondary verification cannot be carried out.

Contents of the invention

In order to overcome the defects in the above-mentioned prior art, the object of the present invention is to provide a method and system for generating an equal-scale real scene map using aerial images. The present invention can use the aerial images to obtain two-dimensional, etc. Scale real scene map, efficiently and accurately restore all elements of the accident scene.

In order to achieve the above-mentioned purpose of the present invention, according to a first aspect of the present invention, the present invention provides a method for generating an equal-scale real-scene scene map using aerial images, comprising the following steps:

S1: Construct an aerial photography platform;

S2: Take a bird's-eye view image of the accident scene;

S3: Carrying out geometric correction to the bird's-eye view image;

S4: Carry out coordinate transformation on the bird's-eye view image, and transform the coordinate system of the bird's-eye view image into an equal-scale real-scene coordinate system.

S5: Use the known reference distance as a scale to extract and mark the accident element information in the equal-scale real scene coordinate system.

The method for generating an equal-scale real-scene scene map by using aerial images of the present invention can use aerial images to obtain a real and detailed two-dimensional image of the accident scene through geometric correction and coordinate transformation, and then generate a detailed traffic accident scene map through image measurement and labeling, which is highly efficient. Accurately restore the scene of the accident. Compared with the existing technology, the measurement and drawing time of the present invention is short and the efficiency is extremely high, which is beneficial to the rapid recovery of the traffic accident scene; the present invention can accurately mark the positions of vehicles, roads, and traces, the key parameters are completely preserved, and the measurement accuracy is high. The measurement error caused by human factors can be avoided and the obtained scene map cannot be modified subjectively; in addition, the present invention can store the image of the accident scene and call it when needed, and can perform secondary verification on the accident scene.

In a preferred embodiment of the present invention, the step of geometric correction is:

S21: Use a two-dimensional calibration board to calibrate the camera, construct an image mapping equation, and obtain internal parameters and external parameters of the camera, the internal parameters including distortion coefficients;

S22: adjust the value of the distortion coefficient to 0, keep other parameters unchanged, and obtain a new internal parameter matrix of the camera;

S23: Substituting the old internal parameters of the camera into the image mapping equation to solve the imaging plane coordinates corresponding to the pixel coordinates of the distorted image, and then using the new camera internal parameters into the image mapping equation to bring the imaging plane coordinate plane into the new image mapping Equation to obtain the corrected image pixel coordinates, and then achieve color restoration through grayscale interpolation.

In another preferred embodiment of the present invention, the steps of using a two-dimensional calibration board to calibrate the camera, constructing an image mapping equation, and obtaining the internal parameters and external parameters of the camera are as follows:

S31: Prepare a two-dimensional calibration plate, the two-dimensional calibration plate has at least two patterns with a certain area that have a high contrast with the background and have a fixed radius and row-column spacing;

S32: Use the camera to be calibrated to take images containing the calibration plate from different angles;

S33: Using an image recognition algorithm to search and locate the pattern in the captured image and extract the centroid coordinates;

S34: Using the two-dimensional actual coordinates of the pattern on the calibration board and the extracted centroid coordinates to establish a mapping equation, seek its optimal solution, and obtain the internal parameters and external parameters of the camera, the mapping equation is:

P _w =P _c *R+Τ;

${<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; f</span>}}{{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; ;</span>$

${<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; u</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; v</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\sqrt{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 4</span>}\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; u</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; v</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span>}}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; ;</span>$

${<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; r</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; c</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; (</span>\frac{{\mathrm{<span\; class="notranslate">\; v</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}{{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; ,</span>\frac{{\mathrm{<span\; class="notranslate">\; u</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}{{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; ;</span>$

Among them, R is a rotation matrix, Τ is a translation matrix, and the R and Τ are external parameters of the camera; P _w is the world coordinate, and (x _w , y _w , z _w ) represent a coordinate point in the world coordinate, P _c For the coordinates transformed into the camera coordinate system, use (x _c , y _c , z _c ) to represent the coordinate points in the camera coordinate system; f is the focal length, u and v are the ideal imaging plane coordinates; k is the distortion coefficient, u′, v′ are the coordinates of the actual imaging plane; r is the number of pixel rows, c is the number of pixel columns, S _x , S _y are the pixel coordinates of the image center, C _x , C _y are the main points in the imaging coordinate system coordinate.

The invention can correct the distortion of the image taken by the camera, obtain the real picture of the accident scene, and restore the scene picture of the accident efficiently and accurately.

In a preferred embodiment of the present invention, the coordinate transformation equation adopted in the step S4 is:

$\left(\begin{array}{c}{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\\ {\mathrm{<span\; class="notranslate">\; the\; y</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\end{array}\right)<span\; class="notranslate">\; =</span>\frac{\left(\begin{array}{cc}\mathrm{<span\; class="notranslate">\; a</span>}& \mathrm{<span\; class="notranslate">\; b</span>}\\ \mathrm{<span\; class="notranslate">\; c</span>}& \mathrm{<span\; class="notranslate">\; d</span>}\end{array}\right)\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}\\ \mathrm{<span\; class="notranslate">\; the\; y</span>}\end{array}\right)<span\; class="notranslate">\; +</span>\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; <figure-callout\; id="1"\; label="e\; f"\; filenames="HDA00002997317300011.png"\; state="\{\{state\}\}">e</figure-callout></span><figure-callout\; id="1"\; label="e\; f"\; filenames="HDA00002997317300011.png"\; state="\{\{state\}\}">\; </figure-callout>}& \\ \mathrm{<span\; class="notranslate">\; </span>}\mathrm{<span\; class="notranslate">f</span>}\end{array}\right)}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\left(\begin{array}{cc}\mathrm{<span\; class="notranslate">\; u</span>}& \mathrm{<span\; class="notranslate">\; v</span>}\end{array}\right)\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}\\ \mathrm{<span\; class="notranslate">\; the\; y</span>}\end{array}\right)}$

Among them, a, b, c, d, e, f, u, v are coordinate conversion parameters. During calculation, 4 reference points are selected to establish 8 equations, and each reference point needs to have two-dimensional ground coordinates and corresponding The two-dimensional image coordinates of .

The invention obtains a real and detailed two-dimensional image of the accident scene through coordinate transformation, and then generates a detailed traffic accident scene map through image annotation.

In order to achieve the above-mentioned purpose of the present invention, according to the second aspect of the present invention, the present invention provides a system for generating an equal-scale real scene map using aerial images, including a mobile device that can be lifted into the air and a camera carried by it. Taking images of the accident scene and transmitting the images to a processor, the processor corrects the images according to the method of the present invention to obtain a real accident scene.

The system of the present invention for generating a digitized scene map by using an aerial image can correct a distorted image taken by a camera, obtain a real picture of an accident scene, and efficiently and accurately restore the picture of the scene of the accident.

In a preferred embodiment of the present invention, the camera transmits the image of the accident scene to the processor through the zigbee module.

Additional aspects and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.

Description of drawings

The above and/or additional aspects and advantages of the present invention will become apparent and comprehensible from the description of the embodiments in conjunction with the following drawings, wherein:

Fig. 1 is the structural representation of the system of the present invention utilizes aerial photography image to generate equal-scale real-scene scene map;

Fig. 2 is a flow chart of the method for generating an equal-scale real-scene scene map by using aerial images in the present invention.

Detailed ways

Embodiments of the present invention are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference numerals designate the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the figures are exemplary only for explaining the present invention and should not be construed as limiting the present invention.

In the description of the present invention, unless otherwise specified and limited, it should be noted that the terms "installation", "connection" and "connection" should be understood in a broad sense, for example, it can be mechanical connection or electrical connection, or two The internal communication of each element may be directly connected or indirectly connected through an intermediary. Those skilled in the art can understand the specific meanings of the above terms according to specific situations.

The present invention provides a system for generating an equal-scale real scene map using aerial images, as shown in Figure 1, which includes a lift-off mobile device 2 and a camera 1 carried therein. Can be but not limited to remote control aircraft. The camera 1 takes images of the accident scene 3 and transmits the images to the processor 4, and the processor corrects the images to obtain a real accident scene. In this embodiment, the camera 1 transmits the images of the accident scene to the processor 4 through the zigbee module, and the processor 4 can process the images captured by the camera in real time. It needs to be said that the shooting method and processing method are the same as when the camera is shooting, and the present invention will not repeat them here.

The present invention also provides a method for generating an equal-scale real-scene scene map using aerial images, as shown in Figure 2, comprising the following steps:

Step 1: Construct an aerial photography platform. In this embodiment, the aerial photography platform is capable of hovering stably and flying with GPS positioning, and includes a mobile device 2 that can be lifted into the air and a camera 1 that it carries.

The second step: shoot the accident scene overlooking image, in this embodiment, can use a certain fixed scene of the accident scene, such as traffic signs, vehicles, etc. At least 4 pattern marking plates may be arranged in the center, the mutual distances of the pattern marking plates are known, and they are used for camera measurement scale calibration. In this embodiment, the shape of the pattern on the pattern sign board is circular. In this embodiment, the mutual distance between the pattern calibration plates is related to the accuracy of coordinate transformation. The larger the distance between the pattern calibration plates, the higher the accuracy of coordinate transformation. In this embodiment, the preferred pattern calibration plate relative Neighbor distance is 2 meters. The invention utilizes the pattern mark board as a coordinate ruler to carry out coordinate transformation on images taken at different heights, which is efficient and accurate.

Step 3: Carry out geometric correction to the top view image. In this embodiment, the camera adopts a wide-angle lens, and the radial distortion of the image is relatively serious. Due to the image measurement, it is necessary to perform geometric correction on the distorted image, use the calibration board and the calibration program to calibrate the aerial camera, and use the update camera internal The post-parameter image mapping equation performs geometric correction on the image. In this embodiment, the steps of geometric correction are:

S31: Calibrate the camera by using a two-dimensional calibration board, construct an image mapping equation, and obtain internal parameters and external parameters of the camera, where the internal parameters include a distortion coefficient. In this embodiment, the specific steps are:

First, a two-dimensional calibration plate is prepared, and the two-dimensional calibration plate has at least 9 patterns with a certain area that have a high contrast with the background, a radius, and a fixed distance between rows and columns. In a preferred embodiment of the present invention, the patterns include each other Separated at least 9 circular graphs.

Then, use the camera to be calibrated to take images containing the calibration plate at different angles, and ensure that the light is sufficient and the shooting pattern is clear when shooting.

Then, use the image recognition algorithm to search and locate the pattern in the captured image and extract the centroid coordinates.

Finally, use the two-dimensional actual coordinates of the pattern on the calibration plate and the extracted centroid coordinates to establish a mapping equation to find its optimal solution. In this embodiment, the least square method is used to solve the mapping equation to obtain the internal parameters and external parameters of the camera. , in this embodiment, the internal parameters include the focal length, the distortion coefficient, the coordinates of the principal point of the camera in the imaging coordinate system, the pixel coordinates of the image center, and the image size; the external parameters include the rotation matrix and translation matrix of the image, so as to complete the camera calibration. In a preferred embodiment of the present invention, the mapping equation adopted is:

P _w =P _c *R+Τ;

${<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; f</span>}}{{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; ;</span>$

${<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; u</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; v</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\sqrt{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 4</span>}\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="u"\; filenames="HDA00002997317300011.png"\; state="\{\{state\}\}">u</figure-callout></span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; v</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span>}}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; ;</span>$

${<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; r</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; c</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; (</span>\frac{{\mathrm{<span\; class="notranslate">\; v</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}{{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; ,</span>\frac{{\mathrm{<span\; class="notranslate">\; u</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}{{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; ;</span>$

Among them, R is a rotation matrix, Τ is a translation matrix, and the R and Τ are external parameters of the camera; P _w is the world coordinate, and (x _w , y _w , z _w ) represent a coordinate point in the world coordinate, P _c For the coordinates transformed into the camera coordinate system, use (x _c , y _c , z _c ) to represent the coordinate points in the camera coordinate system; f is the focal length, u and v are the ideal imaging plane coordinates; k is the distortion coefficient, u′, v′ are the coordinates of the actual imaging plane; r is the number of pixel rows, c is the number of pixel columns, S _x , S _y are the pixel coordinates of the image center, C _x , C _y are the main points in the imaging coordinate system coordinate.

S32: adjust the value of the distortion coefficient to 0, keep other parameters unchanged, and obtain a new internal parameter matrix of the camera;

S33: Substituting the old internal parameters of the camera into the image mapping equation to solve the imaging plane coordinates corresponding to the pixel coordinates of the distorted image, and then using the new camera internal parameters to substitute into the image mapping equation to bring the imaging plane coordinate plane into the new image mapping Equation to obtain the corrected image pixel coordinates, and then achieve color restoration through grayscale interpolation. In this embodiment, the grayscale interpolation method may be, but not limited to, the nearest neighbor method, that is, the grayscale of the nearest neighboring pixel among the four neighboring pixels around the distorted pixel is taken as the grayscale of the point.

The method for generating an equal-scale real-scene scene map by using the aerial photographed image of the present invention can correct the distortion of the image captured by the camera, obtain a real accident scene picture, and restore the accident scene picture efficiently and accurately.

Step 4: Carry out coordinate transformation on the bird's-eye view image, and transform the coordinate system of the bird's-eye view image into an equal-scale real-scene coordinate system.

In this embodiment, homogeneous coordinate transformation is used to transform the captured bird's-eye view image into an equal-scale real-scene coordinate system. The equal-scale real-scene coordinate system has a translation relationship with the ground coordinate system, and the image pixel size of an object on the ground plane is proportional to the actual size. Proportional relationship. In this embodiment, the homogeneous coordinate transformation equation adopted is:

$\left(\begin{array}{c}{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\\ {\mathrm{<span\; class="notranslate">\; the\; y</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\end{array}\right)<span\; class="notranslate">\; =</span>\frac{\left(\begin{array}{cc}\mathrm{<span\; class="notranslate">\; a</span>}& \mathrm{<span\; class="notranslate">\; b</span>}\\ \mathrm{<span\; class="notranslate">\; c</span>}& \mathrm{<span\; class="notranslate">\; d</span>}\end{array}\right)\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}\\ \mathrm{<span\; class="notranslate">\; the\; y</span>}\end{array}\right)<span\; class="notranslate">\; +</span>\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; e</span>}\\ \mathrm{<span\; class="notranslate">\; <figure-callout\; id="1"\; label="f"\; filenames="HDA00002997317300011.png"\; state="\{\{state\}\}">f</figure-callout></span>}\end{array}\right)}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; +</span>\left(\begin{array}{cc}\mathrm{<span\; class="notranslate">\; u</span>}& \mathrm{<span\; class="notranslate">\; v</span>}\end{array}\right)\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}\\ \mathrm{<span\; class="notranslate">\; the\; y</span>}\end{array}\right)}$

Wherein, a, b, c, d, e, f, u, v are the coordinate transformation parameters, when calculating, select 4 reference points from the said pattern mark plate to set up 8 equations, and each reference point needs It has two-dimensional ground coordinates and corresponding two-dimensional image coordinates. In this embodiment, the four reference points can be located in the same pattern mark plate, or in different pattern mark plates, preferably in different pattern mark plates. In a more preferred embodiment of the present invention, each pattern mark Pick a datum point in the board.

In this embodiment, pixel filling may be used to map pixels of the image before coordinate transformation back to the image after coordinate transformation one at a time and determine its gray level. If the pixel of the image before the coordinate transformation is mapped to the four pixels of the image after the coordinate transformation, its gray value is determined by the gray level interpolation algorithm, and the nearest neighbor method is preferably used, that is, four neighboring pixels around the pixel point are taken The grayscale of the nearest neighbor pixel is used as the grayscale of the point, so that the image is more realistic.

Step 5: Using the known reference distance as a scale to extract and mark accident element information in the equal-scale real-scene coordinate system. In this embodiment, the relative distance of the image in the pattern sign board can be used as a coordinate scale to extract and mark the accident element information; it is also possible to select a physical object with a fixed relative position from the accident scene as a coordinate scale to extract and mark the accident element information .

It should be noted that the present invention is not only applicable to the scene of a traffic accident, but also applicable to scenes such as a blasting scene. The method for generating a specific equal-scale real-scene scene map can refer to the steps of the present invention, and these methods are all within the protection scope of the present invention. middle.

The present invention can use the aerial photographed image to obtain a real and detailed two-dimensional image of the accident scene through geometric correction and coordinate transformation, and then generate a detailed traffic accident scene map through image annotation, so as to efficiently and accurately restore the accident scene picture. The measurement and drawing time of the present invention is short and the efficiency is extremely high, which is beneficial to the rapid recovery of the traffic accident scene; the present invention can accurately mark the positions of vehicles, roads, and traces, the key parameters are completely preserved, the measurement accuracy is high, and the measurement caused by human factors can be avoided In addition, the present invention can store the images of the accident scene, call them when needed, and perform secondary verification on the accident scene.

In the description of this specification, descriptions referring to the terms "one embodiment", "some embodiments", "example", "specific examples", or "some examples" mean that specific features described in connection with the embodiment or example , structure, material or characteristic is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

Although the embodiments of the present invention have been shown and described, those skilled in the art can understand that various changes, modifications, substitutions and modifications can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the invention is defined by the claims and their equivalents.

Claims (6)

1. the utilization image of taking photo by plane generates the method for equal proportion outdoor scene scene photo, it is characterized in that, comprises the steps:

S1: construct the platform of taking photo by plane;

S2: take scene of the accident overhead view image;

S3: described overhead view image is carried out geometry correction;

S4: described overhead view image is carried out coordinate transform, with the coordinate system transformation of described overhead view image to equal proportion outdoor scene coordinate system;

S5: in described equal proportion outdoor scene coordinate system, utilize the known reference distance as scale the accident element information to be extracted and mark.

2. the utilization as claimed in claim 1 image of taking photo by plane generates the method for equal proportion outdoor scene scene photo, it is characterized in that the step of described geometry correction is:

S21: utilize two-dimentional scaling board that camera is demarcated, the construct image mapping equation obtains inner parameter and the external parameter of camera, and described inner parameter comprises distortion factor;

S22: the numerical value of distortion factor is adjusted to 0, keeps other parameter constant, obtain the inner parameter matrix of new camera;

S23: utilize the old inner parameter substitution image mapped equation of camera, solve the imaging plane coordinate of distorted image pixel coordinate correspondence, the described new camera inner parameter substitution image mapped equation of recycling, bring the imaging plane coordinate surface into new image mapped equation, image pixel coordinate after obtaining proofreading and correct is realized the color reduction by gray-level interpolation again.

3. the utilization as claimed in claim 2 image of taking photo by plane generates the method for equal proportion outdoor scene scene photo, it is characterized in that the two-dimentional scaling board of described utilization is demarcated camera, and the construct image mapping equation obtains the inner parameter of camera and the step of external parameter and is:

S31: prepare two-dimentional scaling board, have at least 9 patterns of fixing with background contrasts height, radius and ranks spacing with certain area on the described two-dimentional scaling board;

S32: utilize camera to be calibrated to take the image that contains described scaling board of different angles;

S33: utilize image recognition algorithm the pattern in the photographic images to be searched for and located and extract center-of-mass coordinate;

S34: utilize the two-dimentional actual coordinate of pattern on the scaling board and the described center-of-mass coordinate of extraction to set up mapping equation, ask its optimum solution, obtain camera inner parameter and external parameter, described mapping equation is:

P _w=P _c*R+Τ；

${(u,v)}^T=\frac{f}{z_c}{(x_c,y_c)}^T;$

${(u^{\&prime;},v^{\&prime;})}^T=\frac{2}{1+\sqrt{1-4k(u^2+v^2)}}{(u,v)}^T;$

${(r,c)}^T={(\frac{v^{\&prime;}}{S_y}+C_y,\frac{u^{\&prime;}}{S_x}+C_x)}^T;$

Wherein, R is rotation matrix, and Τ is translation matrix, and described R, Τ are the external parameter of camera; P _wBe world coordinates, with (x _w, y _w, z _w) the interior coordinate points of expression world coordinates, P _cFor it transforms to coordinate in the camera coordinates system, with (x _c, y _c, z _c) coordinate points in the expression camera coordinates system; F is focal length, and u, v are desirable imaging plane coordinate; K is distortion factor, and u ', v ' are the actual imaging planimetric coordinates; R is the pixel line number, and c is the pixel columns, S _x, S _yBe picture centre pixel coordinate, C _x, C _yBe the coordinate of principal point in imaging coordinate system.

4. the utilization as claimed in claim 1 image of taking photo by plane generates the method for equal proportion outdoor scene scene photo, it is characterized in that the coordinate transformation equation that adopts among the described step S4 is:

$\left(\begin{array}{c}{x}^{\&prime;}\\ y^{\&prime;}\end{array}\right)=\frac{\left(\begin{array}{cc}a& b\\ c& d\end{array}\right)\left(\begin{array}{c}x\\ y\end{array}\right)+\left(\begin{array}{c}e\\ f\end{array}\right)}{1+\left(\begin{array}{cc}u& v\end{array}\right)\left(\begin{array}{c}x\\ y\end{array}\right)}$

Wherein, a, b, c, d, e, f, u, v are coordinate transformation parameter, when calculating, choose 4 reference points and set up 8 equations, and each reference point all needs to have areal coordinate two-dimensionally and corresponding two dimensional image coordinate.

5. the utilization image of taking photo by plane generates the system of equal proportion outdoor scene scene photo, it is characterized in that, comprise the camera (1) that can go up to the air mobile device (2) and carry, described camera (1) is taken the image of the scene of the accident (3) and described image is transferred to processor (4), described processor (4) the method utilization according to claim 1 image of taking photo by plane generates the digitizing scene photo, obtains real accident scene.

6. the utilization as claimed in claim 5 image of taking photo by plane generates the system of equal proportion outdoor scene scene photo, it is characterized in that, described camera (1) is transferred to processor (4) by the zigbee module with the image of the scene of the accident.

Images