CN105333818B

CN105333818B - 3d space measuring method based on monocular-camera

Info

Publication number: CN105333818B
Application number: CN201410339869.9A
Authority: CN
Inventors: 吴朝晖
Original assignee: Zhejiang Uniview Technologies Co Ltd
Current assignee: Zhejiang Uniview Technologies Co Ltd
Priority date: 2014-07-16
Filing date: 2014-07-16
Publication date: 2018-03-23
Anticipated expiration: 2034-07-16
Also published as: CN105333818A

Abstract

The present invention provides a kind of 3d space measuring method based on monocular-camera, including：The tested point on ground and subtest point are focused respectively, obtain corresponding imaging parameters, wherein, subtest point is located at camera optical axis on the projection line on ground, and tested point is different from subtest point object distance, and the video camera installation parameter being imaged twice is constant；According to tested point and the imaging parameters of subtest point, calculating is associated to being imaged twice, obtain the relative coordinate of tested point, the reference frame of relative coordinate is, origin is projected as on ground with video camera installation fulcrum, using the fulcrum and the vertical line on ground as Y-axis, X-axis is projected as along ground with camera optical axis, using the direction vertical with X-axis and Y-axis as Z axis.The present invention calculates the space coordinates of tested point using the inner parameter of monocular-camera, without measuring the installation parameter of video camera and demarcation thing being demarcated, saves cost, simplifies operation.

Description

3D space measuring method based on monocular camera

Technical Field

The invention relates to the field of video monitoring, in particular to a 3D space measuring method based on a monocular camera.

Background

The monocular camera cannot form 3D vision by a single shot. When the auxiliary measurement is carried out on a reference object without a pre-known size and the erection height of the camera and the included angle between the central axis of the lens and the ground are not known, the relative coordinate between the shot object and the installation point of the camera cannot be measured.

In the prior art, when a monocular camera shoots a single time, a proportional relation between an object size and a pixel of an image shot by the camera is calculated by shooting a reference object with a preset size according to the pixel occupied by the shot reference object in the image, the actual object size and an included angle between the camera and the ground. Then, in the subsequent shooting, the actual size of the object is calculated according to the number of pixels occupied by the shot object without changing the installation parameters of the camera. The relative coordinate between the shot object and the camera can be calculated through the included angle between the camera and the ground and the height of the erection rod.

Through the process, the monocular camera needs to have a plurality of external conditions for single shooting, and needs camera installation data such as the height of an erection rod and the included angle between the camera and the ground; a calibration object is needed to be used, and auxiliary calibration calculation is carried out manually to obtain conversion parameters; the installation parameters of the camera cannot be changed in subsequent use, otherwise the camera must be calibrated again, and the adaptability is poor.

Or the same object is shot by a binocular camera in the prior art, or the same object is shot by a monocular camera at two different positions and angles, and 3D visual perception is realized through different positions in an image. However, it is necessary to use two cameras or to provide a single camera with a moving guide and a driving device, which is costly.

Disclosure of Invention

In view of the above, the present invention provides a 3D space measuring method based on a monocular camera, including:

focusing a point to be tested and an auxiliary test point on the ground respectively to obtain corresponding imaging parameters, wherein the auxiliary test point is positioned on a projection line of an optical axis of a camera on the ground, the object distance between the point to be tested and the auxiliary test point is different, and the camera installation parameters of two times of imaging are unchanged;

and performing correlation calculation on the two imaging according to the imaging parameters of the point to be measured and the auxiliary test point to obtain a relative coordinate of the point to be measured, wherein a reference coordinate system of the relative coordinate is that the projection of a camera mounting fulcrum on the ground is taken as an origin, the perpendicular line of the fulcrum and the ground is taken as a Y axis, the projection of the camera optical axis along the ground is taken as an X axis, and the direction perpendicular to the X axis and the Y axis is taken as a Z axis.

The invention also provides a 3D space measuring device based on the monocular camera, which comprises:

the imaging parameter acquisition unit is used for respectively focusing a point to be measured and an auxiliary test point on the ground to acquire corresponding imaging parameters, wherein the auxiliary test point is positioned on a projection line of a camera optical axis on the ground, the object distance between the point to be measured and the auxiliary test point is different, and the camera installation parameters of the two times of imaging are unchanged;

and the relative coordinate calculation unit is used for performing correlation calculation on the two imaging according to the imaging parameters of the point to be measured and the auxiliary test point to obtain the relative coordinate of the point to be measured, wherein the reference coordinate system of the relative coordinate is that the projection of the camera installation fulcrum on the ground is taken as an origin, the perpendicular line of the fulcrum and the ground is taken as a Y axis, the projection of the camera optical axis along the ground is taken as an X axis, and the direction perpendicular to the X axis and the Y axis is taken as a Z axis.

The invention uses the internal parameters of the monocular camera to calculate the space relative coordinate of the point to be measured, does not need to measure the installation parameters of the camera and calibrate the calibration object, saves the labor, material and time costs and simplifies the operation process.

Drawings

Fig. 1 is a schematic diagram of a logic structure of monocular camera-based 3D spatial measurement and its underlying hardware environment in one embodiment of the present invention.

Fig. 2 is a flowchart of a monocular camera-based 3D spatial measurement method in one embodiment of the present invention.

Fig. 3 is a schematic view of a monocular camera mount.

FIG. 4 is a schematic illustration of optical imaging in one embodiment of the present invention.

FIG. 5 is a schematic diagram of the image height of an imaging point in the Y-axis direction of an image sensor according to an embodiment of the present invention.

Fig. 6 is a schematic view of the principle of lens imaging.

FIG. 7 is a schematic diagram of the image height of an imaging point in the Z-axis direction of an image sensor according to an embodiment of the present invention.

Detailed Description

The present invention will be described in detail below with reference to the accompanying drawings.

The present invention provides a monocular camera based 3D spatial measuring device, which is described below by taking a software implementation as an example, but the present invention does not exclude other implementations such as hardware or logic devices. As shown in fig. 1, the hardware environment in which the apparatus operates includes a CPU, a memory, a nonvolatile memory, and other hardware. The device acts as a logical level virtual device that is run by the CPU. The apparatus includes an imaging parameter acquisition unit and a relative coordinate calculation unit. Referring to fig. 2, the operation of the apparatus includes the following steps:

step 101, an imaging parameter obtaining unit respectively focuses on a point to be tested and an auxiliary test point on the ground to obtain corresponding imaging parameters, wherein the auxiliary test point is located on a projection line of a camera optical axis on the ground, the object distance between the point to be tested and the auxiliary test point is different, and the camera installation parameters of two times of imaging are unchanged;

102, performing correlation calculation on the two imaging according to the imaging parameters of the point to be measured and the auxiliary test point by a relative coordinate calculation unit to obtain a relative coordinate of the point to be measured, wherein a reference coordinate system of the relative coordinate is that the projection of a camera installation fulcrum on the ground is taken as an origin, the perpendicular line of the fulcrum and the ground is taken as a Y axis, the projection of a camera optical axis along the ground is taken as an X axis, and the direction perpendicular to the X axis and the Y axis is taken as a Z axis.

The method images the point to be measured and the auxiliary test point under the condition of not changing the installation parameters (position, height, holder angle and the like) of the monocular camera, and performs correlation calculation on the two images according to the imaging parameters to obtain the spatial coordinate of the point to be measured. The specific processing procedure is as follows.

As shown in fig. 3, the monocular camera is vertically mounted at point E by a pole. And establishing a reference coordinate system by taking the point E as an origin, and calculating the position coordinates of the point to be measured relative to the coordinate system. The X axis of the coordinate system is the projection of the optical axis direction of the camera on the ground, the Y axis is the vertical rod direction, and the Z axis is vertical to the XY plane. In the figure, the intersection point A of an object AD and the ground is a point to be measured, and the point B on the ground is an auxiliary test point, and is projected on the optical axis of the camera along the ground.

As shown in fig. 4, a simple structure of the camera is shown, wherein the optical center of the lens is a virtual optical center formed by multiple lenses of the camera lens, the distance (along the optical axis direction of the camera) between the optical center of the lens and the installation pivot of the camera is r, and the imaging of the two times can be changed. Is the included angle between the optical axis of the camera and the ground.

Under the condition of not changing the installation parameters (height, optical axis angle and direction) of the camera, focusing the point A and the point B respectively by using a monocular camera to obtain corresponding imaging parameters. The imaging point of the point A on the image sensor is a point a, the imaging point of the point B on the image sensor is a point B, and images and objects are projected onto a plane vertical to the optical axis, so that calculation and derivation are facilitated. P₁An object plane for primary imaging, namely an object plane where the point A is located; p₂The object plane of the secondary imaging, namely the object plane where the point B is located. Thereby obtaining the image distance V of one-time imaging₁Focal length F₁And the distance r between the optical center of the lens and the mounting pivot₁(ii) a Image distance V of secondary imaging₂Focal length F₂And the distance r between the optical center of the lens and the mounting pivot₂。

The image height is calculated from the position of the imaging point in the image sensor, as shown in fig. 5. The upper image point in the figure is point b, and the lower image point is point a. And calculating the physical size (image height) of the imaging point along the XY plane according to the direct proportion relation between the physical size and the number of the corresponding pixel points along the direction of the physical size. S in the figure is the physical size of the effective pixel range of the image sensor along the XY plane; s₁The vertical distance from the point a to the central horizontal line of the image sensor is the imaging height of the point A; s₂The vertical distance between the B point and the horizontal line at the center of the image sensor, namely the imaging height of the B point.

Through the process, the image distance V corresponding to the A point is obtained₁Focal length F₁Image height S₁And the distance r between the optical center of the lens and the mounting pivot₁(ii) a Obtaining the image distance V corresponding to the B point₂Focal length F₂Image height S₂And the distance r between the optical center of the lens and the mounting pivot₂. And performing correlation calculation on the two imaging according to the imaging parameters to obtain the relative coordinates of the point A to be measured. The calculation process is described in detail below with reference to fig. 4.

According to the Gaussian imaging formula

Respectively calculating the object distance U of the point A₁And object distance U of point B₂

The distance (in the direction of the optical axis) between the object planes from which the two images are obtained is thus:

according to the imaging principle of the lens shown in FIG. 6

The object heights of points A and B relative to the optical axis of the camera, i.e., the values of n and k in FIG. 4, are determined

From the geometrical relationships in FIG. 4, it can be derived

Substituting the formula (4), the formula (6) and the formula (7) into the formula (8) to obtain

Similarly, it can be derived from the geometric relationship

Substituting the formula (4), the formula (6) and the formula (7) into the formula (10) to obtain

Therefore, the temperature of the molten metal is controlled,

l is the X-axis coordinate of the point A to be measured and is marked as A_x。

Point A is in object plane P₁The upper distance from the plane of the optical axis perpendicular to the ground is A_zThe distance between the image point a of the point A and the plane of which the optical axis is vertical to the ground is a_zThen, then

Similarly, the physical dimension (image height a) of the imaging point a along the Z-axis direction is calculated according to the direct proportion relation between the physical dimension and the number of the corresponding pixel points along the physical dimension direction_z). In fig. 7, Q is the physical size of the effective pixel range of the image sensor along the Z-axis direction.

Substituting the formula (2) into the formula (13) to obtain

A_zThe Z-axis coordinate of the point A to be measured is taken as the coordinate of the Z-axis of the point A to be measured; y-axis coordinate A of point A_yIs 0. Coordinates of point A (A)_x，A_y，A_z) The coordinates of the point a are relative to the coordinates of the preset coordinate system of the point E, and if the physical coordinates of the point E are known, the actual geographic position coordinates of the point a can be calculated by adding the coordinates of the point E to the calculated relative coordinates of the point a.

Therefore, the position of the point to be measured can be measured by performing correlation calculation on the imaging parameters through two or more times of imaging of the monocular camera and by utilizing the internal data of the monocular camera, such as information of image distance, focal length, image sensor size and the like. In the process, the installation parameters of the camera do not need to be measured and the calibration object does not need to be calibrated, so that the labor, material and time costs are saved, and the operation process is simplified.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims

1. A3D space measurement method based on a monocular camera is characterized by comprising the following steps:

2. The method of claim 1, wherein:

x-axis coordinate A in the relative coordinate of the point to be measured_xComprises the following steps:

wherein,

V₁the image distance of the point to be measured is taken as the image distance of the point to be measured;

F₁the focal length of the point to be measured is taken as the focal length of the point to be measured;

S₁the image height of the point to be measured along the XY plane is taken as the image height;

r₁the distance between the optical center of the lens of the point to be measured and the installation fulcrum of the camera along the optical axis direction of the camera;

V₂the image distance of the auxiliary test point is used as the image distance of the auxiliary test point;

F₂the focal distance of the auxiliary test point is;

S₂the image height of the auxiliary test point along the XY plane is taken as the image height;

r₂the distance between the optical center of the lens of the auxiliary test point and the installation fulcrum of the camera along the optical axis direction of the camera.

3. The method of claim 1, wherein:

z-axis coordinate A in the relative coordinate of the point to be measured_zComprises the following steps:

wherein,

a_zand the image height of the point to be measured along the Z-axis direction is obtained.

4. The method of claim 2, wherein:

a is described_xThe specific calculation process is as follows:

the object distance U of the point to be measured₁And the object distance U of the auxiliary test point₂Respectively as follows:

thereby obtaining object planes P imaged twice₁And P₂The distance j + m between the two cameras along the optical axis direction of the camera is as follows:

respectively solving the object height n of the point to be tested, which is vertical to the optical axis of the camera, and the object height k of the auxiliary test point, which is vertical to the optical axis of the camera:

respectively substituting the above parameters into a geometric formulaAnd can obtain the product

A_x＝L₁+L₂。

5. A monocular camera-based 3D spatial measuring device, comprising:

6. The apparatus of claim 5, wherein:

the relative coordinate calculation unit calculates the X-axis coordinate A of the point to be measured_xComprises the following steps:

wherein,

S₁the image height of the point to be measured along the XY plane；

F₂the focal distance of the auxiliary test point is;

7. The apparatus of claim 5, wherein:

the relative coordinate calculation unit calculates the Z-axis coordinate A of the point to be measured_zComprises the following steps:

wherein,

8. The apparatus of claim 6, wherein:

the relative coordinate calculation unit calculates the A_xThe specific process comprises the following steps:

the point to be measuredObject distance U₁And the object distance U of the auxiliary test point₂Respectively as follows:

A_x＝L₁+L₂。