CN109327656A

CN109327656A - The image-pickup method and system of full-view image

Info

Publication number: CN109327656A
Application number: CN201710648161.5A
Authority: CN
Inventors: 刘春�; 张书航; 周源; 曾勇
Original assignee: Shanghai Same Exploration Engineering Technology Co Ltd
Current assignee: Shanghai Same Exploration Engineering Technology Co Ltd
Priority date: 2017-08-01
Filing date: 2017-08-01
Publication date: 2019-02-12

Abstract

The invention discloses a kind of image-pickup method of full-view image and system, described image acquisition method includes the image acquisition step of camera pitching and/or horizontal direction, and the image acquisition step of camera pitch orientation is the following steps are included: S₁₁, with camera pitch angle be β₁The first pitching image is shot, is β with camera pitch angle₂And identical camera horizontal angle shoots the second pitching image when with the first pitching image of shooting；S₁₂, calculate the first degree of overlapping of the second pitching image on the first pitching image；S₁₃, judge whether first degree of overlapping is greater than or equal to a degree of overlapping threshold value, if so, enabling camera subject pitch angleFor β₂, camera withImage is shot, if it is not, by β₂‑β₀Assignment is in β₂, β₀Reduce angle, and return step S for first₁₁.The present invention improves splicing success rate and Image Acquisition efficiency, and present invention can apply to the full-view image shooting fields of unmanned plane.

Description

The image-pickup method and system of full-view image

Technical field

The present invention relates to full-view image shooting field, in particular to the image-pickup method and system of a kind of full-view image.

Background technique

Full-view image generally comprises the image of 360 degree of camera view ambient level, all scenery of vertical 180 degree, panorama shadow As being widely used in monitoring, security protection, virtual reality due to the features such as it contains much information, the true and viewing experience of immersion The fields such as browsing.With the extensive use of unmanned plane, and the full-view image that obtains in the sky of unmanned plane and ground full-view image phase Than having the characteristics that visual angle is big, unobstructed, result of browse is good, having obtained every profession and trade and widely paid close attention to application.

Currently, the Image Acquisition mode of common full-view image is holder camera rotation mode.Holder camera rotation mode It is that a series of images is obtained, to be spliced into full-view image by the rotating camera in camera pitching and horizontal direction.Holder phase Machine has the vertical field of view angle of pitch orientation and the horizontal field of view angle of horizontal direction.The rotation of manual operation holder camera carries out complete It needs to guarantee to reach between single image certain degree of overlapping when the Image Acquisition of scape image, can just be spliced into completely, without gaps and omissions Full-view image, some devices that can automatically snap full-view image also need manually to preset shooting angle.

But the generally existing following defect of the Image Acquisition mode of full-view image in the prior art:

1, splicing success rate is not can guarantee.When the rotation of manual operation camera carries out the Image Acquisition of full-view image, it is difficult to The degree of overlapping between image is held, causes to reduce splicing success rate.For example, cited paper 1 (Chinese invention patent application, publication number: CN104881049A, publication date: 2015.09.02) although disclosed in holder camera can be rotated up and down and move so that All directions image is shot, but needs manually to be operated, therefore not can guarantee splicing success rate.Meanwhile being applied to nobody When machine aviation full-view image, due to being limited by unmanned plane endurance, manual operation is also difficult to ensure the speed that image obtains, from And the Image Acquisition efficiency of unmanned plane aviation full-view image is limited, reduce splicing success rate.

2, preferable Image Acquisition efficiency is not can guarantee.In the prior art by the full-view image of rotating camera from cardon As although acquisition device and system can be with automatic collection image and splicing panorama image, each exposure angles of single image (i.e. the pitch angle and horizontal angle of camera), which still needs, to be manually empirically determined, therefore can not be in the premise for guaranteeing splicing success rate Under reach highest Image Acquisition efficiency, and it is weak to the suitability of camera.For example, cited paper 2 (Chinese invention patent application, Publication number: CN103034042A, publication date: 2013.04.10) and cited paper 3 (Chinese invention patent application, publication number: CN104320581A, publication date: 2015.01.28) in individually disclose panorama shooting method, be required to manually set each bat The position (i.e. exposure angle) of exposure point, therefore Image Acquisition low efficiency are taken the photograph, and diversified camera can not be adapted to, is cited File 4 (U.S. patent Nos application, publication number: US2007263995A1, publication date: 2007.11.15) and (beauty of cited paper 5 State's application for a patent for invention, publication number: US2012294549A1, publication date: 2012.11.22) in individually disclose by carry transport The full-view image shooting method of dynamic detection sensor, but the full-view image resolution ratio generated is fixed, if the track of rotation is not advised Then, the quality that splicing influences can be seriously affected, not can guarantee preferable Image Acquisition efficiency equally.

3, suitable degree of overlapping threshold value can not be determined.In the prior art to degree of overlapping the considerations of, is incomplete, full-view image Need to will splice after image projection to spherical surface, shooting angle overlapping can not accurate description full-view image degree of overlapping, and None suitable degree of overlapping threshold value, (document title: Capturing and Viewing Gigapixel of cited paper 6 Images, author: J Kopf, M Uyttendaele, O Deussen, MF Cohen, open year: 2007) experiment in Using 16% degree of overlapping threshold value, cited paper 7 (document title: panoramic picture obtains and display technology research [D], author: how The ancient cooking vessel universe, Pla Information Engineering University, open year: 2013) in propose at least to need 30% degree of overlapping threshold Value, therefore a unified standard is not provided in the prior art, and is not provided and how to be determined suitable degree of overlapping threshold value Method.

Summary of the invention

The technical problem to be solved by the present invention is to the splicing success rates and image in order to overcome full-view image in the prior art The low defect of collecting efficiency provides the image-pickup method and system of a kind of full-view image.

The present invention is to solve above-mentioned technical problem by following technical proposals:

A kind of image-pickup method of full-view image, it is characterized in that, described image acquisition method includes camera pitching side To image acquisition step, the image acquisition step of the camera pitch orientation the following steps are included:

S₁₁, with camera pitch angle be β₁The first pitching image is shot, is β with camera pitch angle₂And with shooting described first Identical camera horizontal angle shoots the second pitching image when pitching image；

S₁₂, calculate the first degree of overlapping of the second pitching image on the first pitching image；

S₁₃, judge whether first degree of overlapping is greater than or equal to a degree of overlapping threshold value, the degree of overlapping threshold value is used for table The degree of overlapping of splicing is realized between sign image, if so, enabling camera subject pitch angleFor β₂, camera withImage is acquired, and Terminate process, if it is not, reducing β₂, and return step S₁₁；

And/or

Described image acquisition method includes the image acquisition step of camera horizontal direction, the image of the camera horizontal direction Acquisition step the following steps are included:

S₂₁, with camera horizontal angle be α₁First level image is shot, is α with camera horizontal angle₂And with shooting described first Identical camera pitch angle shoots the second horizontal image when horizontal image；

S₂₂, calculate the second degree of overlapping of second horizontal image on the first level image；

S₂₃, judge whether second degree of overlapping is greater than or equal to a degree of overlapping threshold value, the degree of overlapping threshold value is used for table The degree of overlapping of splicing is realized between sign image, if so, enabling camera subject horizontal angle Φ_iFor α₂, camera is with Φ_iImage is acquired, and Terminate process, if it is not, reducing α₂, and return step S₂₁。

In the present solution, when described image acquisition method includes the image acquisition step of camera pitch orientation, by according to Secondary reduction β₂Mode determine the camera pitch angle for meeting default degree of overlapping threshold value, to make phase when panning image Machine automatically determines out ideal Image Acquisition mode in the pitch direction, and pitching is arranged without manual control camera or by rule of thumb Each exposure angle in direction, therefore improve and also improve Image Acquisition efficiency while splicing success rate.

In the present solution, when described image acquisition method includes the image acquisition step of camera horizontal direction, by according to Secondary reduction α₂Mode determine the camera subject horizontal angle for meeting default degree of overlapping threshold value, thus when panning image So that camera is automatically determined out ideal Image Acquisition mode in the horizontal direction, is arranged without manual control camera or by rule of thumb Each exposure angle of horizontal direction, therefore improve and also improve Image Acquisition efficiency while splicing success rate.

In the present solution, when described image acquisition method includes the image acquisition step and camera level of camera horizontal direction It when the image acquisition step in direction, automatically determine camera can in pitch orientation and horizontal direction Ideal Image Acquisition mode out also improves Image Acquisition efficiency while to greatly improve splicing success rate.

Preferably, the camera is bowed when described image acquisition method includes the image acquisition step of camera pitch orientation The image acquisition step for facing upward direction further includes step S₁₄；

In step S₁₃In, camera withAfter acquiring image, step S is executed₁₄；

S₁₄, judgementWhether a pitch angle threshold value is greater than, if so, camera stops acquisition, if it is not, willAssignment is in β₁, and And return step S₁₁。

In the present solution, working asWhen beyond the pitch angle threshold value, camera stops acquisition, bows to be automatically performed in camera Face upward the Image Acquisition on direction.WhenWhen without departing from the pitch angle threshold value, the Image Acquisition in expression camera pitch orientation is still It does not complete, at this time willAssignment is in β₁, and return and continue Image Acquisition.

Preferably, when described image acquisition method includes the image acquisition step of camera pitch orientation, in step S₁₁In, When shooting the second pitching image of starting, β₂≥β₁+ VFOV, VFOV are the vertical field of view for shooting the camera of the full-view image Angle；

In step S₁₃In, if it is not, by β₂-β₀Assignment is in β₂, β₀Reduce angle, and return step S for first₁₁。

Preferably, in step S₁₁In, when shooting originates the second pitching image, β₂=β₁+ VFOV, and/or, in step S₁₃ In, β₀It is 1 degree.

In the present solution, as starting β₂=β₁When+VFOV, it is believed that be the minimum β that the first degree of overlapping is almost 0₂, therefore If originating β₂=β₁+ VFOV, to save calculation amount.

In the present solution, theoretically β₀Smaller outputIt is more accurate, but will increase calculation amount if too small, to place It manages unit and generates burden, and the minimum control precision of pitch orientation of unmanned plane is generally at 1 degree at present, therefore when applied to unmanned plane Preferably 1 degree.

Preferably, when described image acquisition method includes the image acquisition step of camera horizontal direction, in step S₂₃In, If so, enabling camera subject horizontal angle Φ_iFor α₂, camera is with Φ_iImage is acquired, and enables the horizontal angular spacing μ of camera subject_i=Φ_i- α₁, camera is with μ_iAcquire N_iOpen image, N_iNumber, and N are shot for target level_iFor greater than horizontal angle threshold value divided by μ_iMinimum it is whole Number.

It is automatically performed in the present solution, camera is made to pass through fixed target level angular spacing and target level and shoot number Image Acquisition in the horizontal direction.

Preferably, when described image acquisition method includes the image acquisition step of camera horizontal direction, in step S₂₁In, When shooting the second horizontal image of starting, α₂≥α₁+ HFOV, HFOV are the horizontal field of view for shooting the camera of the full-view image Angle；

In step S₁₃In, if it is not, by α₂-α₀Assignment is in α₂, α₀Reduce angle, and return step S for second₂₁。

Preferably, in step S₂₁In, when shooting originates the second horizontal image, α₂=α₁+ HFOV, and/or, in step S₂₃ In, α₀It is 1 degree.

In the present solution, as starting α₂=α₁When+HFOV, it is believed that be the minimum of alpha that the second degree of overlapping is almost 0₂, therefore If originating α₂=α₁+ HFOV, to save calculation amount.

In the present solution, theoretically α₀The μ of smaller output_iIt is more accurate, but if too small, it will increase calculation amount, to processing Unit generates burden, and the minimum control precision of horizontal direction of unmanned plane is generally at 1 degree at present, therefore applied to unmanned plane when is excellent It is selected as 1 degree.

Preferably, described image is adopted when described image acquisition method includes the image acquisition step of camera pitch orientation Set method further includes determining degree of overlapping threshold step, the determining degree of overlapping threshold step the following steps are included:

S₃₁, setting the first degree of overlapping threshold value P₁, and by P₁It is preset as in the image acquisition step of camera pitch orientation Degree of overlapping threshold value；

S₃₂, execute the image acquisition step of camera pitch orientation to acquire pitching image, and by pitching image mosaic be Full-view image；

S₃₃If, splice successfully, reduce P₁, and return step S₃₁If splicing failure, by it is preceding be once spliced into function when P₁ The degree of overlapping threshold value being determined as in the image acquisition step of camera pitch orientation.

In the present solution, the determining degree of overlapping threshold step is to search the process of preferable degree of overlapping threshold value, because existing respectively Class camera and various scenes, therefore there is no an absolutely applicable degree of overlapping threshold value, cause to reduce versatility, but in this programme In, the preferable degree of overlapping threshold value for being suitable for Current camera and scene is determined by the determining degree of overlapping threshold step, because Versatility can be improved in this, while ensure that the quality of full-view image, and suitable degree of overlapping threshold value is smaller, can reduce calculating Amount, therefore Image Acquisition efficiency is improved by determining preferable degree of overlapping threshold value.

Preferably, when described image acquisition method includes the image acquisition step of camera pitching horizontal direction, the figure As acquisition method further includes determining degree of overlapping threshold step, the determining degree of overlapping threshold step the following steps are included:

S₃₁, setting the first degree of overlapping threshold value P₁, and by P₁It is preset as in the image acquisition step of camera horizontal direction Degree of overlapping threshold value；

S₃₂, execute camera horizontal direction image acquisition step to acquire horizontal image, and horizontal image is spliced into Full-view image；

S₃₃If, splice successfully, reduce P₁, and return step S₃₁If splicing failure, by it is preceding be once spliced into function when P₁ The degree of overlapping threshold value being determined as in the image acquisition step of camera horizontal direction.

Preferably, when described image acquisition method includes the image acquisition step and camera horizontal direction of camera pitch orientation Image acquisition step when, described image acquisition method further includes determining degree of overlapping threshold step, the determining degree of overlapping threshold value Step the following steps are included:

S₃₁, setting the first degree of overlapping threshold value P₁, and by P₁It is preset as the image acquisition step and phase of camera pitch orientation Degree of overlapping threshold value in the image acquisition step of machine horizontal direction；

S₃₂, execute the image acquisition step of camera pitch orientation and the image acquisition step of camera horizontal direction is bowed to acquire Image and horizontal image are faced upward, and pitching image and horizontal image are spliced into full-view image；

S₃₃If, splice successfully, reduce P₁, and return step S₃₁If splicing failure, by it is preceding be once spliced into function when P₁ The degree of overlapping threshold value being determined as in the image acquisition step of camera pitch orientation and the image acquisition step of camera horizontal direction.

Preferably, in step S₃₃In, if splicing successfully, by P₁-P₀Assignment is in P₁, P₀For degree of overlapping difference, and return to step Rapid S₃₁If splicing failure, by P₁+P₀It is determined as degree of overlapping threshold value.

Preferably, when described image acquisition method includes the image acquisition step of camera pitch orientation, in step S₁₂In, First degree of overlapping of the second pitching image on the first pitching image is calculated according to image area computation model；

When described image acquisition method includes the image acquisition step of camera horizontal direction, in step S₂₂In, according to institute Image area computation model is stated to calculate second degree of overlapping of second horizontal image on the first level image；

Establish described image areal calculation model the following steps are included:

S₄₁, building image space coordinate system s-x, y, z；

S₄₂, set shooting image phase machine and around y-axis rotate horizontal angle α, and rotate around x axis pitch angle β；

S₄₃, image rotates around x axis to β, and by image projection to panoramic projection face；

S₄₄, image be subjected to α on panoramic projection face translate accordingly；

S₄₅, calculate some subpoint coordinate on corresponding panoramic projection face in image, and according to subpoint coordinate To calculate corresponding area of the image on panoramic projection face.

Existing calculation is that image is rotated around x axis β, is projected later further around y-axis rotation alpha to panoramic projection face, because The calculation formula that this is obtained is extremely complex, in the present solution, by the way that image only to be rotated around x axis to after β projection to panoramic projection The mode in face carrys out simplified formula, to greatly reduce calculation amount, improves the computational efficiency of model.

Preferably, in step S₄₃In, the coordinate (x, y, z) of the point on image is found out into panoramic projection face by following formula On subpoint (x ', y ', z '):

Wherein

A kind of image capturing system of full-view image, it is characterized in that, described image acquisition system includes pitch orientation figure As acquisition module, the pitch orientation image capture module includes first camera control module and first processing module；

The first camera control module is for controlling camera with camera pitch angle as β₁The first pitching image is shot, with phase Machine pitch angle is β₂And the second pitching image is shot with camera horizontal angle identical when shooting the first pitching image, and will The first pitching image and the second pitching image are sent to the first processing module；

After the first processing module is used to receive the first pitching image and the second pitching image, described in calculating First degree of overlapping of the second pitching image on the first pitching image；

The first processing module is also used to judge whether first degree of overlapping is greater than or equal to a degree of overlapping threshold value, institute Degree of overlapping threshold value is stated for characterizing the degree of overlapping for realizing splicing between image, if so, enabling camera subject pitch angleFor β₂, call First camera control module control camera withImage is acquired, if it is not, reducing β₂, and the first camera is called to control Module；

And/or

Described image acquisition system includes horizontal direction image acquisition module, and the horizontal direction image acquisition module includes Second camera control module and Second processing module；

The second camera control module is for controlling camera with camera horizontal angle as α₁First level image is shot, with phase Machine horizontal angle is α₂And the second horizontal image is shot with camera pitch angle identical when shooting the first level image, and will The first level image and second horizontal image are sent to the Second processing module；

After the Second processing module is used to receive the first level image and second horizontal image, described in calculating Second degree of overlapping of second horizontal image on the first level image；

The Second processing module is also used to judge whether second degree of overlapping is greater than or equal to a degree of overlapping threshold value, institute Degree of overlapping threshold value is stated for characterizing the degree of overlapping for realizing splicing between image, if so, enabling camera subject horizontal angle Φ_iFor α₂, call The second camera control module control camera is with Φ_iImage is acquired, if it is not, reducing α₂, and the second camera is called to control Module.

In the present solution, when described image acquisition system includes pitch orientation image capture module, by being sequentially reduced β₂Mode determine the camera pitch angle for meeting default degree of overlapping threshold value, so that camera be made to bow when panning image It faces upward and automatically determines out ideal Image Acquisition mode on direction, pitch orientation is set without manual control camera or by rule of thumb Each exposure angle, therefore improve and also improve Image Acquisition efficiency while splicing success rate.

In the present solution, when described image acquisition system includes horizontal direction image acquisition module, by being sequentially reduced α₂Mode determine the camera subject horizontal angle for meeting default degree of overlapping threshold value, to make camera when panning image Ideal Image Acquisition mode is automatically determined out in the horizontal direction, and level side is set without manual control camera or by rule of thumb To each exposure angle, therefore improve splicing success rate while also improve Image Acquisition efficiency.

In the present solution, when described image acquisition system includes that pitch orientation image capture module and horizontal direction image are adopted When collecting module, make camera that can automatically determine out ideal image in pitch orientation and horizontal direction when panning image Acquisition mode also improves Image Acquisition efficiency while to greatly improve splicing success rate.

Preferably, when described image acquisition system includes pitch orientation image capture module, the first processing module Call first camera control module control camera withAfter shooting image, it is also used to judgeWhether a pitch angle is greater than Threshold value, if so, calling the first camera control module control camera to stop shooting, if it is not, willAssignment is in β₁, and call The first camera control module.

Preferably, when described image acquisition system includes pitch orientation image capture module, shooting the second pitching of starting When image, β₂≥β₁+ VFOV, VFOV are the vertical field of view angle for shooting the camera of the full-view image；

If the first processing module is also used to the first degree of overlapping less than the degree of overlapping threshold value, by β₂-β₀Assignment in β₂, β₀Reduce angle for first, and calls the first camera control module.

Preferably, when shooting the second pitching image of starting, β₂=β₁+VFOV；And/or β₀It is 1 degree.

In the present solution, theoretically β₀Smaller outputIt is more accurate, but will increase calculation amount if too small, to One processing unit generate burden, and at present unmanned plane the minimum control precision of pitch orientation generally at 1 degree, therefore applied to nobody Preferably 1 degree when machine.

Preferably, when described image acquisition system includes horizontal direction image acquisition module, the Second processing module Call the second camera control module control camera with Φ_iAfter acquiring image, it is also used to enable the horizontal angular spacing μ of camera subject_i =α₂-α₁, call the second camera control module control camera with μ_iShoot N_iOpen image, N_iNumber is shot for target level, And N_iFor greater than horizontal angle threshold value divided by μ_iSmallest positive integral.

Preferably, shooting starting second is horizontal when described image acquisition system includes horizontal direction image acquisition module When image, α₂≥α₁+ HFOV, HFOV are the horizontal field of view angle for shooting the camera of the full-view image；

If the Second processing module is also used to the second degree of overlapping less than the degree of overlapping threshold value, by α₂-α₀Assignment in α₂, α₀Reduce angle for second, and calls the second camera control module.

Preferably, when shooting the second horizontal image of starting, α₂=α₁+HFOV；And/or α₀It is 1 degree.

In the present solution, theoretically α₀The μ of smaller output_iIt is more accurate, but if too small, it will increase calculation amount, to second Processing unit generates burden, and the minimum control precision of horizontal direction of unmanned plane is generally at 1 degree at present, therefore applied to unmanned plane When preferably 1 degree.

Preferably, when described image acquisition system includes pitch orientation image capture module, described image acquisition system It further include degree of overlapping threshold determination module；

The degree of overlapping threshold determination module is for setting the first degree of overlapping threshold value P₁, and by P₁It is preset as described in calling Degree of overlapping threshold value when pitch orientation image capture module；

The degree of overlapping threshold determination module is also used to call the pitch orientation image capture module to acquire pitching figure Picture, and be full-view image by pitching image mosaic, if splicing successfully, reduce P₁, and the pitch orientation image is called to adopt Collect module, if splicing failure, by it is preceding be once spliced into function when P₁When being determined as calling the pitch orientation image capture module Degree of overlapping threshold value.

In the present solution, the degree of overlapping threshold determination module is used to search preferable degree of overlapping threshold value, because there are all kinds of phases Machine and various scenes, therefore there is no an absolutely applicable degree of overlapping threshold value, cause to reduce versatility, but in the present embodiment In, the preferable degree of overlapping threshold value for being suitable for Current camera and scene is determined by degree of overlapping threshold determination module, therefore can Versatility is improved, while ensure that the quality of full-view image, and suitable degree of overlapping threshold value is smaller, can reduce calculation amount, because This improves Image Acquisition efficiency by determining preferable degree of overlapping threshold value.

Preferably, when described image acquisition system includes horizontal direction image acquisition module, described image acquisition system It further include degree of overlapping threshold determination module；

The degree of overlapping threshold determination module is for setting the first degree of overlapping threshold value P₁, and by P₁It is preset as described in calling Degree of overlapping threshold value when horizontal direction image acquisition module；

The degree of overlapping threshold determination module is also used to call the horizontal direction image acquisition module to acquire level view Picture, and horizontal image is spliced into full-view image, if splicing successfully, reduce P₁, and the horizontal direction image is called to adopt Collect module, if splicing failure, by it is preceding be once spliced into function when P₁When being determined as calling the horizontal direction image acquisition module Degree of overlapping threshold value.

Preferably, when described image acquisition system includes pitch orientation image capture module and horizontal direction image acquisition mould When block, described image acquisition system further includes degree of overlapping threshold determination module；

The degree of overlapping threshold determination module is for setting the first degree of overlapping threshold value P₁, and by P₁It is preset as described in calling Degree of overlapping threshold value when pitch orientation image capture module and the horizontal direction image acquisition module；

The degree of overlapping threshold determination module is also used to call the pitch orientation image capture module and the level side To image capture module to acquire pitching image and horizontal image, and pitching image and horizontal image are spliced into panorama shadow Picture reduces P if splicing successfully₁, and the pitch orientation image capture module and the horizontal direction image is called to acquire mould Block, if splicing failure, by it is preceding be once spliced into function when P₁It is determined as calling the pitch orientation image capture module and described Degree of overlapping threshold value when horizontal direction image acquisition module.

Preferably, if the degree of overlapping threshold determination module is also used to splice successfully, by P₁-P₀Assignment is in P₁, P₀For overlapping Difference is spent, and by P₁Again it is preset as degree of overlapping threshold value, if splicing failure, by P₁+P₀It is determined as degree of overlapping threshold value.

Preferably, when described image acquisition system includes pitch orientation image capture module, the first processing module For calculating first overlapping of the second pitching image on the first pitching image according to image area computation model Degree；

When described image acquisition system includes horizontal direction image acquisition module, the Second processing module is used for basis Described image areal calculation model calculates second degree of overlapping of second horizontal image on the first level image；

The component of establishing of described image areal calculation model includes:

First establishes component, for constructing image space coordinate system s-x, y, z；

Second establishes component, rotates horizontal angle α around y-axis for setting shooting image phase machine, and rotate around x axis pitch angle β；

Third establishes component, for image to be rotated around x axis β, and by image projection to panoramic projection face；

4th establishes component, translates accordingly for image to be carried out to α on panoramic projection face；

5th establishes component, for calculating the subpoint coordinate on a little corresponding panoramic projection face in image, and Corresponding area of the image on panoramic projection face is calculated according to subpoint coordinate.

Preferably, the coordinate (x, y, z) of the point on image to be found out by following formula to the subpoint on panoramic projection face (x ', y ', z '):

Wherein

On the basis of common knowledge of the art, above-mentioned each optimum condition, can any combination to get each preferable reality of the present invention Example.

The positive effect of the present invention is that:

The present invention realizes full-view image by way of exporting suitable camera pitch angle and camera horizontal angle automatically The automation of Image Acquisition determines each exposure angle without manual control camera or by rule of thumb, is spliced into improve Image Acquisition efficiency is also improved while power.

The present invention also determines the degree of overlapping threshold value for meeting the situation of presence by searching for the mode of ideal degree of overlapping threshold value, Preferable scheme can be provided according to different scene and camera type.

It present invention can be suitably applied to various cameras, it is versatile, the limitation to acquisition platform performance is reduced, to reduce For the equipment cost for the image that pans, make user and obtain with lower equipment cost the image data of preferable quality.

Detailed description of the invention

Fig. 1 is the Image Acquisition step of camera pitch orientation in the image-pickup method of the full-view image of the embodiment of the present invention 1 Rapid flow chart.

Fig. 2 is the Image Acquisition step of camera horizontal direction in the image-pickup method of the full-view image of the embodiment of the present invention 2 Rapid flow chart.

Fig. 3 is the Image Acquisition step of camera pitch orientation in the image-pickup method of the full-view image of the embodiment of the present invention 3 Rapid flow chart.

Fig. 4 is the Image Acquisition step of camera horizontal direction in the image-pickup method of the full-view image of the embodiment of the present invention 3 Rapid flow chart.

Fig. 5 is the process that degree of overlapping threshold step is determined in the image-pickup method of the full-view image of the embodiment of the present invention 3 Figure.

Fig. 6 establishes image area computation model step in the image-pickup method for the full-view image of the embodiment of the present invention 3 Flow chart.

Fig. 7 is the image space coordinate system of the embodiment of the present invention 3 and the schematic diagram of image projection.

Fig. 8 is the schematic diagram that degree of overlapping of the invention defines.

Fig. 9 is the structural schematic diagram of the image capturing system of the full-view image of the embodiment of the present invention 4.

Figure 10 is the structural schematic diagram of the image capturing system of the full-view image of the embodiment of the present invention 5.

Figure 11 is the structural schematic diagram of the image capturing system of the full-view image of the embodiment of the present invention 6.

Specific embodiment

The present invention is further illustrated below by the mode of embodiment, but does not therefore limit the present invention to the reality It applies among a range.

Embodiment 1

As shown in Figure 1, embodiment 1 provides a kind of image-pickup method of full-view image, described image acquisition method includes The image acquisition step of camera pitch orientation, the image acquisition step of the camera pitch orientation the following steps are included:

Step 101 with camera pitch angle is β₁The first pitching image is shot, is β with camera pitch angle₂And with shooting first Identical camera horizontal angle shoots the second pitching image when pitching image.

In this step, β₂≥β₁+ VFOV, VFOV are the vertical field of view angle for shooting the camera of the full-view image, VFOV is decided by the relevant parameter of camera itself.

Step 102 calculates first degree of overlapping of the second pitching image on the first pitching image.

In this step, when the changing value of camera pitch angle be greater than or equal to VFOV when, it is believed that the first pitching image with It is non-overlapping between second pitching image, therefore originate first degree of overlapping of the second pitching image on the first pitching image and be almost 0.To generate error when avoiding and calculating the first degree of overlapping, the first pitching image and phase when the second pitching image of shooting need to will be shot Machine horizontal angle is adjusted to identical.

In this step, the definition of degree of overlapping can refer to shown in Fig. 8, wherein 100,300 be respectively that image A, B are thrown in panorama The projection in shadow face, 200 be degree of overlapping of the image B on image A, is defined as formula 5.

Formula 5:

Wherein, S (Image A) refers to that the area of image A, S (Image A ∩ Image B) refer to that image A is overlapping with image B Area.

Step 103 judges whether first degree of overlapping is greater than or equal to a degree of overlapping threshold value, if so, executing step 104, if it is not, executing step 105.

In this step, the degree of overlapping threshold value is for characterizing the degree of overlapping for realizing splicing between image.

Step 104 enables camera subject pitch angleFor β₂, camera withAcquire image.

Step 105, by β₂-β₀Assignment is in β₂.After executing step 105, return step 101.

In this step, β₀Reduce angle for first, does not limit β specifically in the present embodiment₀Numerical value, can be according to reality Situation is adjusted correspondingly.

In the present embodiment, from starting the first degree of overlapping be almost 0 β₂Start, by successively reducing β₂Mode come it is true The camera pitch angle for meeting default degree of overlapping threshold value is made, to keep camera automatic in the pitch direction when panning image It determines ideal Image Acquisition mode, each explosure angle of pitch orientation is set without manual control camera or by rule of thumb Degree, therefore improve and also improve Image Acquisition efficiency while splicing success rate.

Embodiment 2

As shown in Fig. 2, embodiment 2 provides a kind of image-pickup method of full-view image, described image acquisition method includes The image acquisition step of camera horizontal direction, the image acquisition step of the camera horizontal direction the following steps are included:

Step 201 with camera horizontal angle is α₁First level image is shot, is α with camera horizontal angle₂And with shooting first Identical camera pitch angle shoots the second horizontal image when horizontal image.

In this step, α₂≥α₁+ HFOV, HFOV are the horizontal field of view angle for shooting the camera of the full-view image, HFOV is decided by the relevant parameter of camera itself.

Step 202 calculates second degree of overlapping of second horizontal image on first level image.

In this step, when the changing value of camera horizontal angle be greater than or equal to HFOV when, it is believed that first level image with It is non-overlapping between second horizontal image, therefore originate second degree of overlapping of second horizontal image on first level image and be almost 0.To generate error when avoiding and calculating the second degree of overlapping, first level image and phase when the second horizontal image of shooting need to will be shot Machine pitch angle is adjusted to identical.

Formula 5:

Step 203 judges whether second degree of overlapping is greater than or equal to a degree of overlapping threshold value, if so, executing step 204, if it is not, executing step 205.

Step 204 enables camera subject horizontal angle Φ_iFor α₂, camera is with Φ_iAcquire image.

Step 205, by α₂-α₀Assignment is in α₂.After executing step 205, return step 201.

In this step, α₀Reduce angle for second, does not limit α specifically in the present embodiment₀Numerical value, can be according to reality Situation is adjusted correspondingly.

In the present embodiment, from starting the second degree of overlapping be almost 0 α₂Start, by successively reducing α₂Mode come it is true The camera horizontal angle for meeting default degree of overlapping threshold value is made, to keep camera automatic in the horizontal direction when panning image It determines ideal Image Acquisition mode, each explosure angle of horizontal direction is set without manual control camera or by rule of thumb Degree, therefore improve and also improve Image Acquisition efficiency while splicing success rate.

Embodiment 3

Embodiment 3 provides a kind of image-pickup method of full-view image, and described image acquisition method includes camera pitching side To image acquisition step, the image acquisition step of camera horizontal direction and determine degree of overlapping threshold step, in the present embodiment, Described image acquisition method is applied in the full-view image shooting of unmanned plane, is especially applied to single by rotating in same place Platform camera is come in the unmanned plane for the image that pans, it is not limited to this application scenarios, can also be applied to mobile terminal or list The full-view image shooting etc. of reverse phase machine can be selected according to user demand.

Specifically, as shown in figure 3, the image acquisition step of the camera pitch orientation the following steps are included:

In this step, β₂=β₁+ VFOV, VFOV are the vertical field of view angle for shooting the camera of the full-view image, VFOV is decided by the relevant parameter of camera itself.

In this step, the degree of overlapping threshold value is for characterizing the degree of overlapping for realizing splicing between image, the degree of overlapping Threshold users are preset in combination with practical scene and experience, it is proposed that it is first preset as biggish numerical value, it in the present embodiment, will be described Degree of overlapping threshold preset is 30%, it is not limited to this numerical value.

Step 104 enables camera subject pitch angleFor β₂, camera withAcquire image.

In this step, it is produced in cyclic process multiple and differentValue, can successively be denoted as

In this step, β₀Reduce angle for first, in the present embodiment β₀It is 1 degree, theoretically β₀Smaller outputMore Accurately, it will increase calculation amount but if too small, burden generated to processing unit, and the pitch orientation of unmanned plane is minimum at present Precision is controlled generally at 1 degree, therefore is preferably 1 degree in the present embodiment, but do not limit β specifically₀Numerical value, can be according to practical feelings Condition is adjusted correspondingly.

In the present embodiment, from starting the first degree of overlapping be almost 0 β₂Start, by successively reducing β₂Mode come it is true The camera pitch angle for meeting default degree of overlapping threshold value is made, as starting β₂=β₁When+VFOV, it is believed that be the first degree of overlapping almost For 0 minimum β₂, therefore set starting β₂=β₁+ VFOV, so that calculation amount is saved, it is not limited to this numerical value, can also set Beginning β₂>β₁+ VFOV can be selected according to the actual situation.

Step 106, judgementWhether pitch angle threshold value is greater than, if so, step 107 is executed, if it is not, executing step 108.

In this step, it is contemplated that the limitation of unmanned plane hardware, the pitch angle threshold value are set as 90 degree, it is not limited to This numerical value can also be adjusted correspondingly according to the actual situation.

Step 107, camera stop acquisition.

In this step, whenWhen beyond the pitch angle threshold value, camera stops shooting, bows to be automatically performed in camera Face upward the Image Acquisition on direction.

Step 108 is incited somebody to actionAssignment is in β₁.After executing step 108, return step 101.

In this step, whenWhen without departing from the pitch angle threshold value, the Image Acquisition in expression camera pitch orientation is still It does not complete, at this time willAssignment is in β₁, and step 101 is back to continue Image Acquisition.In the present embodiment, panorama Image is therefore each exported adjacent output by will splice after image projection to spherical surfaceBetween difference not It is identical, thus by from first time outputStart successively to be denoted asAnd it is stored.

As shown in figure 4, the image acquisition step of the camera horizontal direction the following steps are included:

Step 201 is with camera pitch angleCamera horizontal angle is α₁First level image is shot, with camera pitch angle ForCamera horizontal angle is α₂Shoot the second horizontal image.

In this step, α₂=α₁+ HFOV, HFOV are the horizontal field of view angle for shooting the camera of the full-view image, HFOV is decided by the relevant parameter of camera itself.It is determined by the image acquisition step of the camera pitch orientationLater, The image acquisition step for starting to execute the camera horizontal direction often determines one in the present embodimentLater, start to hold The image acquisition step of the row camera horizontal direction, but also can determine that allAnd then start to execute each The image acquisition step of corresponding camera horizontal direction can be selected according to user's actual need.

Step 203 judges whether second degree of overlapping is greater than degree of overlapping threshold value, if so, step 2041 is executed, if it is not, Execute step 205.

In this step, the degree of overlapping threshold value is using the degree of overlapping in the image acquisition step of the camera pitch orientation Threshold value.

Step 2041 enables the horizontal angular spacing μ of camera subject_i=α₂-α₁, camera is with μ_iAcquire N_iOpen image.

In this step, N_iNumber, and N are shot for target level_iFor greater than horizontal angle threshold value divided by μ_iSmallest positive integral, Camera shoots number by fixed target level angular spacing and target level to be automatically performed Image Acquisition in the horizontal direction. In the present embodiment, the horizontal angle threshold value is 360 degree, i.e., rotation is turned around in the horizontal direction, but can also be according to the actual situation To select the angle rotated in the horizontal direction.

In this step, α₀Reduce angle for second, in the present embodiment α₀It is 1 degree, theoretically α₀The μ of smaller output_iMore Accurately, but if it is too small, it will increase calculation amount, processing unit generated and is born, and the minimum control of horizontal direction of unmanned plane at present Precision processed is preferably 1 degree in the present embodiment, but do not limit α specifically generally at 1 degree₀Numerical value, can be according to the actual situation To be adjusted correspondingly.

In the present embodiment, from starting the second degree of overlapping be almost 0 α₂Start, by successively reducing α₂Mode come it is true The horizontal angular spacing of camera for meeting default degree of overlapping threshold value is made, as starting α₂=α₁When+HFOV, it is believed that be the second degree of overlapping Almost 0 minimum of alpha₂, therefore set starting α₂=α₁+ HFOV, can also be with it is not limited to this numerical value to save calculation amount If originating α₂>α₁+ HFOV can be selected according to the actual situation.

As shown in figure 5, the determining degree of overlapping threshold step the following steps are included:

Step 301, the first degree of overlapping threshold value P of input₁, and by P₁It is preset as the image acquisition step of camera pitch orientation And the degree of overlapping threshold value in the image acquisition step of camera horizontal direction.

In this step, the first degree of overlapping threshold value P₁Preset degree of overlapping threshold value is originated for user.

The Image Acquisition step of step 302, the image acquisition step for successively executing camera pitch orientation and camera horizontal direction Suddenly to acquire pitching image and horizontal image, and pitching image and horizontal image are spliced into full-view image.

Step 303 judges whether splicing succeeds, if splicing successfully, executes step 304, if splicing failure, executes step 305。

Step 304, by P₁-P₀Assignment is in P₁.After executing step 304, return step 301.

In this step, P₀For degree of overlapping difference, in the present embodiment, although not limiting P specifically₀Numerical value, but consider Circulation needs huge calculation amount every time, therefore can be by P₀It is set as 5%.

In this step, if splicing successfully, then it represents that the leeway that this degree of overlapping threshold value can also reduce, therefore reduce again Degree of overlapping threshold value is simultaneously recycled.

Step 305, by P₁+P₀The image of the image acquisition step and camera horizontal direction that are determined as camera pitch orientation is adopted Collect the degree of overlapping threshold value in step.

In this step, if splicing failure, then it represents that this degree of overlapping threshold value has reached the limit, therefore by previous degree of overlapping Threshold value, i.e. P₁+P₀It is determined as preferable degree of overlapping threshold value.

In the present embodiment, the determining degree of overlapping threshold step is to search the process of preferable degree of overlapping threshold value, because existing All kinds of cameras and various scenes, therefore there is no an absolutely applicable degree of overlapping threshold value, cause to reduce versatility, but in this reality It applies in example, the preferable degree of overlapping threshold for being suitable for Current camera and scene is determined by the determining degree of overlapping threshold step Value, therefore versatility can be improved, while ensure that the quality of full-view image, and suitable degree of overlapping threshold value is smaller, can reduce Calculation amount, therefore Image Acquisition efficiency is improved by determining preferable degree of overlapping threshold value.

In the step 102 of the present embodiment, the second pitching image is calculated according to image area computation model in institute State the first degree of overlapping on the first pitching image.In the step 202 of the present embodiment, according to described image areal calculation model come Calculate second degree of overlapping of second horizontal image on the first level image.

Detailed description below establishes the step of described image areal calculation model.

As shown in fig. 6, establish described image areal calculation model the following steps are included:

Step 401, building image space coordinate system s-x, y, z.

In this step, establish pixel number that data that image area computation model need to input are camera original image (W, H), focal length f, camera horizontal angle α and camera pitch angle β, wherein α ∈ (- π, π], β ∈ [- pi/2, pi/2].

According to the pixel number (W, H) of original image, wherein W and H is respectively the width and height (pixel) of image, constructs picture Plane coordinate system IJ, then image space coordinate system s-x, y, z are constructed, meet the corresponding relationship of formula 1.

Formula 1:

Step 402 sets shooting image phase machine and around y-axis rotates horizontal angle α, and rotates around x axis pitch angle β.

In this step, as shown in fig. 7, when shooting image, if phase chance rotates horizontal angle α around y-axis and x-axis respectively and bows Elevation angle β.

Image is rotated around x axis β by step 403, and by image projection to panoramic projection face.

In this step, as shown in fig. 7, because projection of shape of the image on panoramic projection face is unrelated with horizontal angle α, institute That image only can be rotated around x axis β, and by image projection to panoramic projection face, A is image before projecting, and B is projection, obtains formula 2, wherein s-x, y, z are the image space coordinate system coordinate of image.

Formula 2:

Wherein

Existing calculation is that image is rotated around x axis β, is projected later further around y-axis rotation alpha to panoramic projection face, because The calculation formula that this is obtained is extremely complex, in the present embodiment, by projecting after image is only rotated around x axis β to panorama throwing The mode in shadow face carrys out simplified formula, to greatly reduce calculation amount, improves the computational efficiency of model.

If θ andThe corresponding horizontal angle of plane and vertical angle where respectively representing an image, θ ∈ (- π, π],Wherein θ is clockwise direction,It is positive downwards.Horizontal angle θ corresponding to one image each pixel with Vertical angleIt is acquired by formula 3.

Formula 3:

Image is carried out α on panoramic projection face and translates accordingly by step 404.

In this step, camera rotates horizontal angle α around y-axis and can be considered as in the translation of the panoramic projection face direction Shang Yanθ, obtains Projection of the photo on panoramic projection face then has θ=θ+α, i.e., by θ+α assignment in θ.

The subpoint coordinate on a little corresponding panoramic projection face in step 405, calculating image, and according to subpoint Coordinate calculates corresponding area of the image on panoramic projection face.

In this step, by the calculating of formula 2, formula 3 and step 405, available field angle is FOV, shooting direction point Not Wei α, β the projection B that with angular way is stored of the image shot by camera on the wide high panoramic projection face for being respectively W and H, so And actual full-view image is stored with pixel coordinate, so need θ andPixel rectangular co-ordinate is converted to by formula 4.

Formula 4:

(the x obtained by formula 4_p,y_p) it is that some subpoint on corresponding panoramic projection face on shooting image is sat Mark, the combination of these coordinates constitutes projection of the filmed image on panoramic projection face, and then can find out the area of projection.

The necessary condition that full-view image is able to splice is that the degree of overlapping between photo must reach a certain degree of overlapping threshold value, this Place's degree of overlapping refers to two photos in the degree of overlapping in panoramic projection face, as shown in figure 8, wherein 100,300 respectively image A, B are complete The projection on scape perspective plane, 200 be degree of overlapping of the image B on image A, is defined as formula 5.

Formula 5:

Wherein, S (Image A) refers to that the area of image A, S (Image A ∩ Image B) refer to that image A is overlapping with image B Area, therefore the first degree of overlapping and the second degree of overlapping can be calculated by described image areal calculation model.

The image-pickup method of full-view image provided in this embodiment by exporting suitable camera pitch angle and phase automatically The mode of machine horizontal angle realizes the automation of the Image Acquisition of full-view image, determines without manual control camera or by rule of thumb Each exposure angle also improves Image Acquisition efficiency while to improve splicing success rate.

The image-pickup method of full-view image provided in this embodiment is also by searching for the mode of ideal degree of overlapping threshold value It determines the degree of overlapping threshold value for meeting the situation of presence, preferable scheme can be provided according to different scene and camera type.

The image-pickup method of full-view image provided in this embodiment is applicable to various cameras, versatile, reduces Limitation to acquisition platform performance makes user with lower equipment to reduce the equipment cost for the image that pans At the image data for obtaining preferable quality originally.

Embodiment 4

As shown in figure 9, embodiment 4 provides a kind of image capturing system 1 of full-view image, the Image Acquisition system of full-view image System 1 includes pitch orientation image capture module 2, and pitch orientation image capture module 2 includes first camera control module 21 and the One processing module 22.

Specifically, first camera control module 21 is for controlling camera with camera pitch angle as β₁Shoot the first pitching figure Picture is β with camera pitch angle₂And identical camera horizontal angle shoots the second pitching image, β when with the first pitching image of shooting₂≥ β₁+ VFOV, VFOV are the vertical field of view angle for shooting the camera of the full-view image, and VFOV is decided by the correlation of camera itself Parameter, and the first pitching image and the second pitching image are sent to first processing module 22.

First processing module 22 is for after receiving the first pitching image and the second pitching image, calculating described the First degree of overlapping of the two pitching images on the first pitching image.

The definition of degree of overlapping can refer to shown in Fig. 8, wherein 100,300 be respectively the projection of image A, B in panoramic projection face, 200 be degree of overlapping of the image B on image A, is defined as formula 5.

Formula 5:

In the present embodiment, when the changing value of camera pitch angle is greater than or equal to VFOV, it is believed that the first pitching image It is non-overlapping between the second pitching image, therefore originate first degree of overlapping of the second pitching image on the first pitching image almost It is 0.For avoid calculate the first degree of overlapping when generate error, need to by shoot the first pitching image and shoot the second pitching image when Camera horizontal angle is adjusted to identical.

First processing module 22 is also used to judge whether first degree of overlapping is greater than or equal to a degree of overlapping threshold value, if It is to enable camera subject pitch angleFor β₂, call first camera control module 21 control camera withImage is acquired, if it is not, will β₂-β₀Assignment is in β₂, β₀Reduce angle for first, and calls first camera control module 21, not specific limit in the present embodiment Determine β₀Numerical value, can be adjusted correspondingly according to the actual situation.In the present embodiment, the degree of overlapping threshold value is for characterizing The degree of overlapping of splicing is realized between image.

Embodiment 5

As shown in Figure 10, embodiment 5 provides a kind of image capturing system 1 of full-view image, the Image Acquisition of full-view image System 1 include horizontal direction image acquisition module 3, horizontal direction image acquisition module 3 include second camera control module 31 and Second processing module 32.

Specifically, second camera control module 31 is for controlling camera with camera horizontal angle as α₁Shoot first level figure Picture is α with camera horizontal angle₂And identical camera pitch angle shoots the second horizontal image, α when with shooting first level image₂≥ α₁+ HFOV, HFOV are the horizontal field of view angle for shooting the camera of the full-view image, and HFOV is decided by the correlation of camera itself Parameter, and the first level image and second horizontal image are sent to Second processing module 32.

Second processing module 32 is for after receiving the first level image and second horizontal image, calculating described the Second degree of overlapping of two horizontal images on the first level image.

Formula 5:

In the present embodiment, when the changing value of camera horizontal angle is greater than or equal to HFOV, it is believed that first level image It is non-overlapping between the second horizontal image, therefore originate second degree of overlapping of second horizontal image on first level image almost It is 0.For avoid calculate the second degree of overlapping when generate error, need to by shoot first level image and shoot the second horizontal image when Camera pitch angle is adjusted to identical.

Second processing module 32 is also used to judge whether second degree of overlapping is greater than or equal to a degree of overlapping threshold value, if It is to enable camera subject horizontal angle Φ_iFor α₂, second camera control module 31 is called to control camera with Φ_iImage is acquired, if it is not, will α₂-α₀Assignment is in α₂, α₀Reduce angle for second, and calls second camera control module 31, not specific limit in the present embodiment Determine α₀Numerical value, can be adjusted correspondingly according to the actual situation.In the present embodiment, the degree of overlapping threshold value is for characterizing The degree of overlapping of splicing is realized between image.

Embodiment 6

As shown in figure 11, embodiment 6 provides a kind of image capturing system 1 of full-view image, the Image Acquisition of full-view image System 1 includes pitch orientation image capture module 2, horizontal direction image acquisition module 3 and degree of overlapping threshold determination module 4, is bowed Facing upward directional image acquisition module 2 includes first camera control module 21 and first processing module 22, and horizontal direction image acquires mould Block 3 includes second camera control module 31 and Second processing module 32, and in the present embodiment, described image acquisition system is set to On unmanned plane, applied in the full-view image shooting of unmanned plane, be especially applied to by same place rotate single camera Come in the unmanned plane for the image that pans, it is not limited to this application scenarios, can also be applied to mobile terminal or slr camera Full-view image shooting etc., can be selected according to user demand.

First camera control module is for controlling camera with camera pitch angle as β₁The first pitching image is shot, is bowed with camera The elevation angle is β₂And the second pitching image is shot with identical camera horizontal angle when the first pitching image of shooting, and by described first Pitching image and the second pitching image are sent to the first processing module, in the present embodiment, β₂=β₁+ VFOV, VFOV For the vertical field of view angle of the camera for shooting the full-view image, VFOV is decided by the relevant parameter of camera itself.

First processing module is for calculating described second after receiving the first pitching image and the second pitching image First degree of overlapping of the pitching image on the first pitching image.In the present embodiment, when the changing value of camera pitch angle is big When VFOV, it is believed that it is non-overlapping between the first pitching image and the second pitching image, therefore originate the second pitching figure As the first degree of overlapping on the first pitching image is almost 0.To generate error when avoiding and calculating the first degree of overlapping, need to will shoot First pitching image is adjusted with camera horizontal angle when the second pitching image of shooting to identical.

First processing module is also used to judge whether first degree of overlapping is greater than or equal to a degree of overlapping threshold value, if so, Enable camera subject pitch angleFor β₂, call first camera control module control camera withAcquire image, recursive call It produces in the process multiple and differentValue, can successively be denoted as If it is not, by β₂-β₀Assignment is in β₂, β₀It is first Reduce angle, and calls the first camera control module.

In the present embodiment, the degree of overlapping threshold value is for characterizing the degree of overlapping for realizing splicing between image, the overlapping Degree threshold users are preset in combination with practical scene and experience, it is proposed that biggish numerical value are first preset as, in the present embodiment, by institute Stating degree of overlapping threshold preset is 30%, it is not limited to this numerical value.

β in the present embodiment₀It is 1 degree, theoretically β₀Smaller outputIt is more accurate, but will increase meter if too small Calculation amount generates burden to first processing units, and the minimum control precision of pitch orientation of unmanned plane is generally at 1 degree at present, therefore It is preferably 1 degree in the present embodiment, but does not limit β specifically₀Numerical value, can be adjusted correspondingly according to the actual situation.

First processing module call first camera control module control camera withAfter shooting image, it is also used to judge Whether a pitch angle threshold value is greater than, if so, calling first camera control module control camera to stop shooting, if it is not, willAssignment In β₁, and call first camera control module.

In the present embodiment, it is contemplated that the limitation of unmanned plane hardware, the pitch angle threshold value is set as 90 degree, but simultaneously not only limits In this numerical value, can also be adjusted correspondingly according to the actual situation.WhenWhen beyond the pitch angle threshold value, camera stops clapping It takes the photograph, to be automatically performed the Image Acquisition in camera pitch orientation.WhenWhen without departing from the pitch angle threshold value, camera is indicated Image Acquisition in pitch orientation is not yet completed, and at this time willAssignment is in β₁, and control camera again to continue image Acquisition.In the present embodiment, full-view image is by will splice after image projection to spherical surface, therefore each exported is adjacent OutputBetween difference be all different, therefore by from first time outputStart successively to be denoted asAnd into Row storage.

Second camera control module is used to control camera with camera pitch angleCamera horizontal angle is α₁Shoot the first water Flat image is with camera pitch angleCamera horizontal angle is α₂Shoot the second horizontal image, and by the first level image And second horizontal image is sent to Second processing module.

In the present embodiment, α₂=α₁+ HFOV, HFOV are the horizontal field of view angle for shooting the camera of the full-view image, HFOV is decided by the relevant parameter of camera itself.It is determined by the pitch orientation image capture moduleLater, start to adjust With horizontal direction image acquisition module, one is often determinedLater, start to call horizontal direction image acquisition module, but can also It determines allAnd then start to call horizontal direction image acquisition module, it can be carried out according to user's actual need Selection.

Second processing module is for calculating described second after receiving the first level image and second horizontal image Second degree of overlapping of the horizontal image on the first level image.In the present embodiment, when the changing value of camera horizontal angle is big When HFOV, it is believed that it is non-overlapping between first level image and the second horizontal image, therefore originate the second level view As the second degree of overlapping on first level image is almost 0.

Second processing module is also used to judge whether second degree of overlapping is greater than the degree of overlapping threshold value, if so, enabling phase Machine target level angular spacing μ_i=α₂-α₁, call second camera control module control camera with μ_iAcquire N_iOpen image, N_iFor target Level shooting number, and N_iFor greater than horizontal angle threshold value divided by μ_iSmallest positive integral, if it is not, by α₂-α₀Assignment is in α₂, α₀It is second Reduce angle, and calls second camera control module.

Camera shoots number by fixed target level angular spacing and target level to be automatically performed in the horizontal direction Image Acquisition.In the present embodiment, the horizontal angle threshold value is 360 degree, i.e., rotation is turned around in the horizontal direction, but can also basis Actual conditions select the angle rotated in the horizontal direction.

α in the present embodiment₀It is 1 degree, theoretically α₀The μ of smaller output_iIt is more accurate, but if too small, it will increase calculating Amount, to the second processing unit generate burden, and at present unmanned plane the minimum control precision of horizontal direction generally at 1 degree, therefore originally It is preferably 1 degree in embodiment, but does not limit α specifically₀Numerical value, can be adjusted correspondingly according to the actual situation.

Degree of overlapping threshold determination module is for receiving the first degree of overlapping threshold value P₁, and by P₁It is preset as calling pitch orientation Degree of overlapping threshold value when image capture module and horizontal direction image acquisition module, in the present embodiment, the first degree of overlapping threshold value P₁Preset degree of overlapping threshold value is originated for user.

Degree of overlapping threshold determination module is also used to that pitch orientation image capture module and horizontal direction image is called to acquire mould Pitching image and horizontal image are spliced into full-view image to acquire pitching image and horizontal image by block, if splicing successfully, By P₁-P₀Assignment is in P₁, P₀For degree of overlapping difference, and pitch orientation image capture module and horizontal direction image is called to acquire Module, if splicing failure, by P₁+P₀When being determined as calling pitch orientation image capture module and horizontal direction image acquisition module Degree of overlapping threshold value, in the present embodiment, although not limiting P specifically₀Numerical value, but in view of circulation every time needs huge meter Calculation amount, therefore can be by P₀It is set as 5%.

In the present embodiment, if splicing successfully, then it represents that the leeway that this degree of overlapping threshold value can also reduce, therefore subtract again Small degree of overlapping threshold value is simultaneously recycled.If splicing failure, then it represents that this degree of overlapping threshold value has reached the limit, therefore will be previous heavy Folded degree threshold value, i.e. P₁+P₀It is determined as preferable degree of overlapping threshold value.

In the present embodiment, degree of overlapping threshold determination module is for searching preferable degree of overlapping threshold value, because there are all kinds of cameras And various scenes, therefore there is no an absolutely applicable degree of overlapping threshold value, cause to reduce versatility, but in the present embodiment, The preferable degree of overlapping threshold value for being suitable for Current camera and scene is determined by degree of overlapping threshold determination module, therefore can be improved Versatility, while ensure that the quality of full-view image, and suitable degree of overlapping threshold value is smaller, calculation amount can be reduced, therefore logical It crosses and determines preferable degree of overlapping threshold value to improve Image Acquisition efficiency.

In the present embodiment, first processing module is for calculating the second pitching figure according to image area computation model As the first degree of overlapping on the first pitching image, Second processing module be used for according to described image areal calculation model come Calculate second degree of overlapping of second horizontal image on the first level image.

Detailed description below described image areal calculation model establishes component.

Described image areal calculation model establish component include first establish component, second establish component, third establish Component, the 4th establish component and the 5th establish component.

Specifically, first component is established for constructing image space coordinate system s-x, y, z.

Establish pixel number (W, H), the focal length f, phase that the data that image area computation model need to input are camera original image Machine horizontal angle α and camera pitch angle β, wherein α ∈ (- π, π], β ∈ [- pi/2, pi/2].

Formula 1:

Second, which establishes component, rotates horizontal angle α around y-axis for setting shooting image phase machine, and rotates around x axis pitch angle β.

As shown in fig. 7, when shooting image, if phase chance rotates horizontal angle α and pitch angle β around y-axis and x-axis respectively.

Third establishes component for image to be rotated around x axis β, and by image projection to panoramic projection face.

As shown in fig. 7, because projection of shape of the image on panoramic projection face is unrelated with horizontal angle α, it can be only by figure As rotating around x axis β, and by image projection to panoramic projection face, A is image before projecting, and B is projection, obtains formula 2, wherein s-x, Y, z are the image space coordinate system coordinate of image.

Formula 2:

In

Formula 3:

4th establishes component translates accordingly for image to be carried out α on panoramic projection face.

Camera, which rotates horizontal angle α around y-axis, can be considered as in the translation of the panoramic projection face direction Shang Yanθ, obtain photo in panorama Projection on perspective plane then has θ=θ+α, i.e., by θ+α assignment in θ.

5th establishes component for calculating the subpoint coordinate on a little corresponding panoramic projection face in image, and root Corresponding area of the image on panoramic projection face is calculated according to subpoint coordinate.

By the calculating of formula 2, formula 3 and step 405, available field angle is FOV, and shooting direction is respectively the phase of α, β Machine shoots the projection B that with angular way stores of the image on the wide high panoramic projection face for being respectively W and H, however actual complete Scape image is stored with pixel coordinate, thus need θ andPixel rectangular co-ordinate is converted to by formula 4.

Formula 4:

Formula 5:

The image capturing system of full-view image provided in this embodiment by exporting suitable camera pitch angle and phase automatically The mode of machine horizontal angle realizes the automation of the Image Acquisition of full-view image, determines without manual control camera or by rule of thumb Each exposure angle also improves Image Acquisition efficiency while to improve splicing success rate.

The image capturing system of full-view image provided in this embodiment is also by searching for the mode of ideal degree of overlapping threshold value It determines the degree of overlapping threshold value for meeting the situation of presence, preferable scheme can be provided according to different scene and camera type.

The image capturing system of full-view image provided in this embodiment may be disposed at various cameras, versatile, reduce Limitation to acquisition platform performance makes user with lower equipment to reduce the equipment cost for the image that pans At the image data for obtaining preferable quality originally.

Although specific embodiments of the present invention have been described above, it will be appreciated by those of skill in the art that this is only For example, protection scope of the present invention is to be defined by the appended claims.Those skilled in the art without departing substantially from Under the premise of the principle and substance of the present invention, many changes and modifications may be made, but these change and Modification each falls within protection scope of the present invention.

Claims

1. a kind of image-pickup method of full-view image, which is characterized in that described image acquisition method includes camera pitch orientation Image acquisition step, the image acquisition step of the camera pitch orientation the following steps are included:

S₁₁, with camera pitch angle be β₁The first pitching image is shot, is β with camera pitch angle₂And with shoot the first pitching figure Identical camera horizontal angle shoots the second pitching image when picture；

S₁₃, judge whether first degree of overlapping is greater than or equal to a degree of overlapping threshold value, the degree of overlapping threshold value is used for phenogram The degree of overlapping of splicing is realized as between, if so, enabling camera subject pitch angleFor β₂, camera withImage is acquired, and is terminated Process, if it is not, reducing β₂, and return step S₁₁；

And/or

Described image acquisition method includes the image acquisition step of camera horizontal direction, the Image Acquisition of the camera horizontal direction Step the following steps are included:

S₂₁, with camera horizontal angle be α₁First level image is shot, is α with camera horizontal angle₂And with shoot the first level figure Identical camera pitch angle shoots the second horizontal image when picture；

S₂₃, judge whether second degree of overlapping is greater than or equal to a degree of overlapping threshold value, the degree of overlapping threshold value is used for phenogram The degree of overlapping of splicing is realized as between, if so, enabling camera subject horizontal angle Φ_iFor α₂, camera is with Φ_iImage is acquired, and is terminated Process, if it is not, reducing α₂, and return step S₂₁。

2. image-pickup method as described in claim 1, which is characterized in that when described image acquisition method includes camera pitching When the image acquisition step in direction, the image acquisition step of the camera pitch orientation further includes step S₁₄；

S₁₄, judgementWhether a pitch angle threshold value is greater than, if so, camera stops acquisition, if it is not, willAssignment is in β₁, and return Return step S₁₁。

3. image-pickup method as described in claim 1, which is characterized in that when described image acquisition method includes camera pitching When the image acquisition step in direction, in step S₁₁In, when shooting originates the second pitching image, β₂≥β₁+ VFOV, VFOV be for Shoot the vertical field of view angle of the camera of the full-view image；

4. image-pickup method as claimed in claim 3, which is characterized in that in step S₁₁In, shooting the second pitching figure of starting When picture, β₂=β₁+ VFOV, and/or, in step S₁₃In, β₀It is 1 degree.

5. image-pickup method as described in claim 1, which is characterized in that when described image acquisition method includes camera level When the image acquisition step in direction, in step S₂₃In, if so, enabling camera subject horizontal angle Φ_iFor α₂, camera is with Φ_iAcquisition figure Picture, and enable the horizontal angular spacing μ of camera subject_i=Φ_i-α₁, camera is with μ_iAcquire N_iOpen image, N_iNumber is shot for target level, And N_iFor greater than horizontal angle threshold value divided by μ_iSmallest positive integral.

6. image-pickup method as described in claim 1, which is characterized in that when described image acquisition method includes camera level When the image acquisition step in direction, in step S₂₁In, when shooting originates the second horizontal image, α₂≥α₁+ HFOV, HFOV be for Shoot the horizontal field of view angle of the camera of the full-view image；

7. image-pickup method as claimed in claim 6, which is characterized in that in step S₂₁In, shooting the second level view of starting When picture, α₂=α₁+ HFOV, and/or, in step S₂₃In, α₀It is 1 degree.

8. image-pickup method as described in claim 1, which is characterized in that when described image acquisition method includes camera pitching When the image acquisition step in direction, described image acquisition method further includes determining degree of overlapping threshold step, the determining degree of overlapping Threshold step the following steps are included:

S₃₁, setting the first degree of overlapping threshold value P₁, and by P₁The degree of overlapping being preset as in the image acquisition step of camera pitch orientation Threshold value；

S₃₂, execute the image acquisition step of camera pitch orientation to acquire pitching image, and be panorama by pitching image mosaic Image；

S₃₃If, splice successfully, reduce P₁, and return step S₃₁If splicing failure, by it is preceding be once spliced into function when P₁It determines For the degree of overlapping threshold value in the image acquisition step of camera pitch orientation.

9. image-pickup method as described in claim 1, which is characterized in that when described image acquisition method includes camera pitching When the image acquisition step of horizontal direction, described image acquisition method further includes determining degree of overlapping threshold step, the determining weight Folded degree threshold step the following steps are included:

S₃₁, setting the first degree of overlapping threshold value P₁, and by P₁The degree of overlapping being preset as in the image acquisition step of camera horizontal direction Threshold value；

S₃₂, execute camera horizontal direction image acquisition step to acquire horizontal image, and horizontal image is spliced into panorama Image；

S₃₃If, splice successfully, reduce P₁, and return step S₃₁If splicing failure, by it is preceding be once spliced into function when P₁It determines For the degree of overlapping threshold value in the image acquisition step of camera horizontal direction.

10. image-pickup method as described in claim 1, which is characterized in that when described image acquisition method includes that camera is bowed Face upward direction image acquisition step and camera horizontal direction image acquisition step when, described image acquisition method further include determine Degree of overlapping threshold step, the determining degree of overlapping threshold step the following steps are included:

S₃₁, setting the first degree of overlapping threshold value P₁, and by P₁Image acquisition step and the camera for being preset as camera pitch orientation are horizontal Degree of overlapping threshold value in the image acquisition step in direction；

S₃₂, the image acquisition step of image acquisition step and camera horizontal direction that executes camera pitch orientation acquire pitching figure Picture and horizontal image, and pitching image and horizontal image are spliced into full-view image；

S₃₃If, splice successfully, reduce P₁, and return step S₃₁If splicing failure, by it is preceding be once spliced into function when P₁It determines For the degree of overlapping threshold value in the image acquisition step of camera pitch orientation and the image acquisition step of camera horizontal direction.

11. the image-pickup method as described in any one of claim 8~10, which is characterized in that in step S₃₃In, if spelling It is connected into function, by P₁-P₀Assignment is in P₁, P₀For degree of overlapping difference, and return step S₃₁If splicing failure, by P₁+P₀Determination is attached most importance to Folded degree threshold value.

12. image-pickup method as described in claim 1, which is characterized in that when described image acquisition method includes that camera is bowed When facing upward the image acquisition step in direction, in step S₁₂In, the second pitching image is calculated according to image area computation model The first degree of overlapping on the first pitching image；

When described image acquisition method includes the image acquisition step of camera horizontal direction, in step S₂₂In, according to the figure Image planes accumulate computation model to calculate second degree of overlapping of second horizontal image on the first level image；

S₄₁, building image space coordinate system s-x, y, z；

S₄₅, calculate some subpoint coordinate on corresponding panoramic projection face in image, and counted according to subpoint coordinate Calculate corresponding area of the image on panoramic projection face.

13. image-pickup method as claimed in claim 12, which is characterized in that in step S₄₃In, by the seat of the point on image Mark (x, y, z) finds out the subpoint (x ', y ', z ') on panoramic projection face by following formula；

Wherein

14. a kind of image capturing system of full-view image, which is characterized in that described image acquisition system includes pitch orientation image Acquisition module, the pitch orientation image capture module include first camera control module and first processing module；

The first camera control module is for controlling camera with camera pitch angle as β₁The first pitching image is shot, is bowed with camera The elevation angle is β₂And the second pitching image is shot with camera horizontal angle identical when shooting the first pitching image, and will be described First pitching image and the second pitching image are sent to the first processing module；

The first processing module is for calculating described second after receiving the first pitching image and the second pitching image First degree of overlapping of the pitching image on the first pitching image；

The first processing module is also used to judge whether first degree of overlapping is greater than or equal to a degree of overlapping threshold value, described heavy Folded degree threshold value is for characterizing the degree of overlapping for realizing splicing between image, if so, enabling camera subject pitch angleFor β₂, described in calling First camera control module control camera withImage is acquired, if it is not, reducing β₂, and call the first camera control mould Block；

And/or

The second camera control module is for controlling camera with camera horizontal angle as α₁First level image is shot, with camera water The straight angle is α₂And the second horizontal image is shot with camera pitch angle identical when shooting the first level image, and will be described First level image and second horizontal image are sent to the Second processing module；

The Second processing module is for calculating described second after receiving the first level image and second horizontal image Second degree of overlapping of the horizontal image on the first level image；

The Second processing module is also used to judge whether second degree of overlapping is greater than or equal to a degree of overlapping threshold value, described heavy Folded degree threshold value is for characterizing the degree of overlapping for realizing splicing between image, if so, enabling camera subject horizontal angle Φ_iFor α₂, described in calling Second camera control module controls camera with Φ_iImage is acquired, if it is not, reducing α₂, and call the second camera control mould Block.

15. image capturing system as claimed in claim 14, which is characterized in that when described image acquisition system includes pitching side When to image capture module, the first processing module call the first camera control module control camera withShoot image Later, it is also used to judgeWhether a pitch angle threshold value is greater than, if so, the first camera control module control camera is called to stop It only shoots, if it is not, willAssignment is in β₁, and call the first camera control module.

16. image capturing system as claimed in claim 14, which is characterized in that when described image acquisition system includes pitching side When to image capture module, when shooting the second pitching image of starting, β₂≥β₁+ VFOV, VFOV are for shooting the full-view image Camera vertical field of view angle；

If the first processing module is also used to the first degree of overlapping less than the degree of overlapping threshold value, by β₂-β₀Assignment is in β₂, β₀For First reduces angle, and calls the first camera control module.

17. image capturing system as claimed in claim 16, which is characterized in that when shooting the second pitching image of starting, β₂=β₁ +VFOV；And/or β₀It is 1 degree.

18. image capturing system as claimed in claim 14, which is characterized in that when described image acquisition system includes level side When to image capture module, the Second processing module calls the second camera control module control camera with Φ_iAcquisition figure As after, it is also used to enable the horizontal angular spacing μ of camera subject_i=α₂-α₁, call second camera control module control camera with μ_iShoot N_iOpen image, N_iNumber, and N are shot for target level_iFor greater than horizontal angle threshold value divided by μ_iSmallest positive integral.

19. image capturing system as claimed in claim 14, which is characterized in that when described image acquisition system includes level side When to image capture module, when shooting the second horizontal image of starting, α₂≥α₁+ HFOV, HFOV are for shooting the full-view image Camera horizontal field of view angle；

If the Second processing module is also used to the second degree of overlapping less than the degree of overlapping threshold value, by α₂-α₀Assignment is in α₂, α₀For Second reduces angle, and calls the second camera control module.

20. image capturing system as claimed in claim 19, which is characterized in that when shooting the second horizontal image of starting, α₂=α₁ +HFOV；And/or α₀It is 1 degree.

21. image capturing system as claimed in claim 14, which is characterized in that when described image acquisition system includes pitching side When to image capture module, described image acquisition system further includes degree of overlapping threshold determination module；

The degree of overlapping threshold determination module is for setting the first degree of overlapping threshold value P₁, and by P₁It is preset as calling the pitching Degree of overlapping threshold value when directional image acquisition module；

The degree of overlapping threshold determination module is also used to call the pitch orientation image capture module to acquire pitching image, and And by pitching image mosaic be full-view image, if splice successfully, reduction P₁, and call the pitch orientation Image Acquisition mould Block, if splicing failure, by it is preceding be once spliced into function when P₁It is determined as calling the weight when pitch orientation image capture module Folded degree threshold value.

22. image capturing system as claimed in claim 14, which is characterized in that when described image acquisition system includes level side When to image capture module, described image acquisition system further includes degree of overlapping threshold determination module；

The degree of overlapping threshold determination module is for setting the first degree of overlapping threshold value P₁, and by P₁It is preset as calling the level Degree of overlapping threshold value when directional image acquisition module；

The degree of overlapping threshold determination module is also used to call the horizontal direction image acquisition module to acquire horizontal image, and And horizontal image is spliced into full-view image, if splicing successfully, reduce P₁, and call the horizontal direction image acquisition mould Block, if splicing failure, by it is preceding be once spliced into function when P₁It is determined as calling the weight when horizontal direction image acquisition module Folded degree threshold value.

23. image capturing system as claimed in claim 14, which is characterized in that when described image acquisition system includes pitching side When to image capture module and horizontal direction image acquisition module, described image acquisition system further includes that degree of overlapping threshold value determines mould Block；

The degree of overlapping threshold determination module is for setting the first degree of overlapping threshold value P₁, and by P₁It is preset as calling the pitching Degree of overlapping threshold value when directional image acquisition module and the horizontal direction image acquisition module；

The degree of overlapping threshold determination module is also used to call the pitch orientation image capture module and the horizontal directivity pattern As acquisition module to acquire pitching image and horizontal image, and pitching image and horizontal image are spliced into full-view image, if Splice successfully, reduces P₁, and the pitch orientation image capture module and the horizontal direction image acquisition module are called, if Splicing failure, by it is preceding be once spliced into function when P₁It is determined as calling the pitch orientation image capture module and the level side Degree of overlapping threshold value when to image capture module.

24. the image capturing system as described in any one of claim 21~23, which is characterized in that the degree of overlapping threshold value If determining module is also used to splice successfully, by P₁-P₀Assignment is in P₁, P₀For degree of overlapping difference, and by P₁Again it is preset as being overlapped Threshold value is spent, if splicing failure, by P₁+P₀It is determined as degree of overlapping threshold value.

25. image capturing system as claimed in claim 14, which is characterized in that when described image acquisition system includes pitching side When to image capture module, the first processing module is for calculating the second pitching figure according to image area computation model As the first degree of overlapping on the first pitching image；

When described image acquisition system includes horizontal direction image acquisition module, the Second processing module is used for according to Image area computation model calculates second degree of overlapping of second horizontal image on the first level image；

5th establishes component, for calculating some subpoint coordinate on corresponding panoramic projection face in image, and according to Subpoint coordinate calculates corresponding area of the image on panoramic projection face.

26. image capturing system as claimed in claim 25, which is characterized in that the coordinate (x, y, z) of the point on image is logical It crosses following formula and finds out subpoint (x ', y ', z ') on panoramic projection face；

Wherein