CN109785278A - A kind of three-dimensional sufficient type image processing method, device, electronic equipment and storage medium - Google Patents
A kind of three-dimensional sufficient type image processing method, device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN109785278A CN109785278A CN201811570138.XA CN201811570138A CN109785278A CN 109785278 A CN109785278 A CN 109785278A CN 201811570138 A CN201811570138 A CN 201811570138A CN 109785278 A CN109785278 A CN 109785278A
- Authority
- CN
- China
- Prior art keywords
- feature point
- point set
- fitting
- error
- obtains
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 20
- 238000003860 storage Methods 0.000 title claims abstract description 17
- 239000011159 matrix material Substances 0.000 claims abstract description 69
- 230000009466 transformation Effects 0.000 claims abstract description 52
- 238000000034 method Methods 0.000 claims abstract description 32
- 230000004927 fusion Effects 0.000 claims abstract description 9
- 238000013519 translation Methods 0.000 claims description 23
- 238000004590 computer program Methods 0.000 claims description 17
- 238000005070 sampling Methods 0.000 claims description 4
- 239000011521 glass Substances 0.000 description 11
- 238000012545 processing Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 6
- 238000005259 measurement Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 239000003638 chemical reducing agent Substances 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000000691 measurement method Methods 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000005304 joining Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 239000005341 toughened glass Substances 0.000 description 1
Landscapes
- Image Processing (AREA)
Abstract
The embodiment of the present invention provides a kind of three-dimensional sufficient type image processing method, device, electronic equipment and storage medium, this method comprises: obtaining point cloud P, Q of adjacent two frames foot type image respectively;The multiple characteristic point informations for obtaining described cloud P form feature point set Pi, the feature point set PiSpace coordinate including the multiple characteristic point;It is obtained on described cloud Q and corresponds to the feature point set PiFeature point set Qi, by transformation matrix by the feature point set QiWith the feature point set PiIt is fitted, obtains fitting result, the fitting result includes error of fitting;If error of fitting is less than preset error threshold, multiple image is iterated according to above step, obtains iteration result;According to the iteration result, the feature point set fusion figure of the multiple image is exported.Since the initial transformation matrix that the point cloud by adjacent two frames foot type image obtains is smaller, the number of point-cloud fitting iteration is reduced, so as to improve the speed of image mosaic.
Description
Technical field
The present invention relates to field of image processing more particularly to a kind of three-dimensional sufficient type image processing methods, device, electronic equipment
And storage medium.
Background technique
With the improvement of living standards, people require also to be continuously improved to the comfort of shoes, health perception also constantly enhances.
Shoes and personal sufficient shape parameter can be made to match by the method that foot-measuring customizes shoes, bring preferably comfortable experience with it is higher
Safety, obtained more and more concerns.
The key of foot-measuring customization is the accurate measurement to personal sufficient shape parameter, and existing foot measurement method can divide manually
It measures and two kinds of three-dimensional scanning measurement.Manual measurement, usually designer take ruler or tape measuring foot type by hand, obtain foot
Long, foot breadth and foot some girth information, this some characteristic point for just needing to find on foot go the ginsengs such as positioning length and width, girth
Number, these characteristic points are demarcated according to the experience of designer, ununified standard, what different designers measured
Foot type parameter is not fully identical, and the shoes for customizing out also have very big difference, so the method low efficiency of this manual measurement
Under, but also customization shoes can not be accomplished to standardize;Three-dimensional scanning measurement generally uses and is based on optical measurement method, at present state
It is inside and outside to have more companies and research institution in progress correlative study, such as the Bostjan Novak of Ljubljana university etc.
Researcher is looped around around foot using 4 pairs of CCD cameras, with the scanned foot of beam of laser line, utilizes laser multi-thread three
Angular measurement is spliced into a complete foot type three-dimensional figure;Hyunglae Lee of Seoul National University et al. uses stereopsis
The method of feel has carried out foot type 3-D scanning, they have used 12 PC cameras to be scanned shooting, passes through Feature Points Matching
The depth that each is put in foot type is estimated, then carries out three-dimensional reconstruction.
Country's foot type 3-D scanning at present, there are also gaps with foreign countries in terms of 3-D image Processing Algorithm.Zhou Jiali et al.
3-D scanning has been carried out to foot type using Binocular Vision Principle, using the digital camera of high definition, has scanned each side of foot
Spliced again after depth map, but more due to counting, calculation amount is very big, and the distance of each iteration is smaller, the speed of splicing
Degree is slower, and splicing is caused to take a long time.
Summary of the invention
The embodiment of the present invention provides a kind of three-dimensional sufficient type image processing method, device, electronic equipment and storage medium, can
Improve the speed of image mosaic.
In a first aspect, the embodiment of the present invention provides a kind of three-dimensional sufficient type image processing method, comprising:
Point cloud P, Q of adjacent two frames foot type image are obtained respectively;
The multiple characteristic point informations for obtaining described cloud P form feature point set Pi, the feature point set PiIncluding the multiple
The space coordinate of characteristic point;
It is obtained on described cloud Q and corresponds to the feature point set PiFeature point set Qi, by transformation matrix by the spy
Levy point set QiWith the feature point set PiIt is fitted, obtains fitting result, the fitting result includes error of fitting;
If error of fitting is less than preset error threshold, multiple image is iterated according to above step,
Obtain iteration result;
According to the iteration result, the feature point set fusion figure of the multiple image is exported.
Optionally, if being less than preset error threshold in the error of fitting, according to above step to multiframe
Before the step for image is iterated, and obtains iteration result, further comprise the steps of:
If error of fitting is greater than the error threshold of setting, continue to obtain on described cloud Q corresponding to the feature
Point set PiFeature point set Qi, and update the transformation matrix;
By updated transformation matrix by the feature point set QiWith the feature point set PiIt is fitted, is fitted
As a result, the fitting result includes error of fitting.
Optionally, the transformation matrix includes spin matrix and translation matrix, and the spin matrix includes rotation initial value,
The translation matrix includes translation initial value.
Optionally, obtaining the rotation initial value according to the angle difference of the adjacent two frames foot type picture point cloud is root
Obtaining the translation initial value according to the distance difference of the adjacent two frames foot type picture point cloud is.
Optionally, described obtain on described cloud Q corresponds to the feature point set PiFeature point set Qi, pass through transformation
Matrix is by the feature point set QiWith the feature point set PiIt is fitted, obtains fitting result, the fitting result includes fitting
Error, comprising:
Feature point set Q is up-sampled in described cloud Qi, so that QiCorresponding to PiThe sum of the distance between characteristic point be less than away from
From threshold value, the distance threshold is calculated by least square method;
By feature point set QiBy transformation matrix rotation translation after with the feature point set PiIt is fitted, obtains the fitting
Error.
Optionally, described to update the transformation matrix, comprising: to be calculated according to the distance threshold and update transformation square
Battle array.
Optionally, if the error of fitting is less than preset error threshold, according to above step to multiframe figure
As being iterated, iteration result is obtained, comprising:
By the transformation matrix by feature point set QiRotation and translation obtains the feature point set QiTo the feature point set
PiMean longitudinal error;
If the mean longitudinal error is less than the preset error threshold, continue to change to next frame image
Generation;
If the mean longitudinal error is greater than the preset error threshold, continue to up-sample in described cloud Q
Corresponding to the feature point set PiFeature point set Qi。
Second aspect, the embodiment of the present invention provide a kind of device of three-dimensional sufficient type image processing method, comprising:
First obtains module, for obtaining point cloud P, Q of adjacent two frames foot type image respectively;
Second obtains module, and multiple characteristic point informations for obtaining described cloud P form feature point set Pi, the feature
Point set PiSpace coordinate including the multiple characteristic point;
Fitting module corresponds to the feature point set P for obtaining on described cloud QiFeature point set Qi, pass through change
Matrix is changed by the feature point set QiWith the feature point set PiIt is fitted, obtains fitting result, the fitting result includes quasi-
Close error;
Judgment module, if being less than preset error threshold for error of fitting, according to above step to multiframe
Image is iterated, and obtains iteration result;
Output module, for exporting the feature point set fusion figure of the multiple image according to the iteration result.
The third aspect, the embodiment of the present invention provide a kind of electronic equipment, comprising: memory, processor and are stored in described
On memory and the computer program that can run on the processor, the processor are realized when executing the computer program
Step in three-dimensional sufficient type image processing method provided in an embodiment of the present invention.
Fourth aspect, the embodiment of the present invention provide a kind of computer readable storage medium, which is characterized in that the computer
It is stored with computer program on readable storage medium storing program for executing, realizes that the embodiment of the present invention mentions when the computer program is executed by processor
Step in the sufficient type image processing method of the three-dimensional of confession.
In the embodiment of the present invention, point cloud P, Q of adjacent two frames foot type image are obtained respectively;Obtain the multiple of described cloud P
Characteristic point information forms feature point set Pi, the feature point set PiSpace coordinate including the multiple characteristic point;In described cloud
It is obtained on Q and corresponds to the feature point set PiFeature point set Qi, by transformation matrix by the feature point set QiWith the feature
Point set PiIt is fitted, obtains fitting result, the fitting result includes error of fitting;It is preset if error of fitting is less than
Error threshold, then multiple image is iterated according to above step, obtains iteration result;It is defeated according to the iteration result
The feature point set of the multiple image merges figure out.The initial transformation square obtained due to the point cloud by adjacent two frames foot type image
Battle array is smaller, the number of point-cloud fitting iteration is reduced, so as to improve the speed of image mosaic.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
The structure shown according to these attached drawings obtains other attached drawings.
Fig. 1 be the embodiment of the invention provides a kind of three-dimensional sufficient type image-scanning device;
Fig. 2 is a kind of flow diagram of three-dimensional sufficient type image processing method provided in an embodiment of the present invention;
Fig. 3 is a kind of flow diagram of specific embodiment in Fig. 1 before step 104;
Fig. 4 is a kind of flow diagram of specific embodiment of step 103 in Fig. 1;
Fig. 5 is the structural schematic diagram of a kind of electronic equipment provided in an embodiment of the present invention;
Fig. 6 is a kind of structural schematic diagram of three-dimensional sufficient type image processing apparatus provided in an embodiment of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are some of the embodiments of the present invention, instead of all the embodiments.Based on this hair
Embodiment in bright, every other implementation obtained by those of ordinary skill in the art without making creative efforts
Example, shall fall within the protection scope of the present invention.
It illustrates under introduction in conjunction with relevant drawings first below, a kind of three-dimensional sufficient type figure that the embodiment of the present invention may be used
As scanning means, as shown in Figure 1, above-mentioned apparatus includes: transparent groove 1, driving device 2, connecting rod rest 3, scanner 4, bracket 5.
Wherein, transparent groove 1 can be made by tempered glass and be formed for placing foot, and be able to bear certain weight
Such as 200 kilograms of amount;The transparent groove 1 includes glass film plates and the glass side plate that the glass film plates two sides are arranged in, and in glass
The joining place of glass bottom plate and glass side plate has carried out designed arc-shaped appearance, so that imaging is indeformable, the image of acquisition is more accurate;Its
In, glass film plates is used to load-bearing and fixed position, and glass side plate can be used to for the glass groove being fixed on the bracket 5;
In addition, being provided with foot-shape pattern on above-mentioned glass film plates, the placement region for prompting biped foot can be used to.
It is provided with mounting groove in the middle part of bracket 5, which can be square slot, U-type groove, for placing above-mentioned transparent groove
1, and limit the position of the transparent groove 1.
Driving device 2 is mounted on the side of bracket 5 and is sequentially connected with 3 one end of connecting rod rest, and scanner 4 is arranged in connecting rod rest
3 other end;Preferably, driving device 2 includes motor, retarder and drive shaft;Wherein, motor can be alternating current generator, use
Power is provided, and has corresponding control circuit and drive system: control signal is sent to motor, to control by control circuit
The rotating forward of motor inverts, and is then rotated by retarder band moving gear, then the driving of the drive shaft by being sequentially connected with said gear
Connecting rod rest 3 rotates back and forth, to drive the scanner 4 for being fixed on the other end on connecting rod rest 3 to rotate back and forth to scan positioned at transparent
Foot in groove 1, and scanner 4 is only needed to rotate the foot 3-D image that can once obtain complete biped, it can be improved
Scan the efficiency of foot image file.
Preferably, connecting rod rest 3 includes: first connecting rod, second connecting rod, cross bar;One end of first connecting rod or second connecting rod can
It is sequentially connected with the drive shaft of above-mentioned driving device 2, to drive the first connecting rod or second connecting rod to turn by driving device
It is dynamic;The both ends of cross bar are provided in the middle part of the cross bar by being detachably connected on the other end of above-mentioned first connecting rod, second connecting rod
Above-mentioned scanner 4, and being rotated together with first connecting rod, second connecting rod, and during rotation scanner 4 camera lens edge always
Diameter be directed toward above-mentioned transparent glass groove, it is ensured that the image scanned is complete.
Optionally, 5 two sides of bracket are respectively arranged with bearing mounting plate and retarder mounting plate, on bearing mounting plate
It is provided with bearing block, bearing is provided on bearing block and is connect with a tailing axle, the tailing axle and above-mentioned first connecting rod or second connecting rod
One end is attached, and the tailing axle is overlapped with the axis of the drive shaft of above-mentioned driving device 2, and above-mentioned scanner 4 can be made to protect
Maintain an equal level and quietly rotates.
Preferably, above-mentioned bracket 5 is equipped with limit switch, and the trigger unit of the limit switch is in above-mentioned first connecting rod or second
On the motion profile of connecting rod;The limit switch can be used to limit the rotating range of first connecting rod and second connecting rod, such as 270 degree, make
Back rotation can be carried out in the range by obtaining first connecting rod and second connecting rod, so that above-mentioned scanner 4, which rotates, can once sweep
Retouch entire foot.
Optionally, stating for above-mentioned driving device is provided with extension spring structure between retarder and retarder mounting plate, the tension spring
It is even closer that structure can make above-mentioned motor gear and reducer gear engage, and gap is smaller, and motor switches from rotating forward
Return difference to two gears during reversion is smaller, and then the operation of connecting rod rest 3 can be made more steady, and scanner 4 collects
Image it is more uniform.
Optionally, above-mentioned scanner 4 is two generation Kinect depth cameras, and high resolution is small by environment shadow sound, can be obtained
Obtain the depth image compared with high definition.
As shown in Fig. 2, the embodiment of the invention provides a kind of three-dimensional sufficient type image processing methods, comprising the following steps:
101, point cloud P, Q of adjacent two frames foot type image are obtained respectively.
Wherein, above-mentioned sufficient type image can be the deep image information by three-dimensional scanner, depth camera acquisition,
The deep image information further includes depth value information in addition to the RGB information including pixel, i.e., where pixel to depth camera
Vertical plane distance value;Above-mentioned cloud is multiple pixel collections with depth value information.
For example, above-mentioned adjacent two can be obtained by the above-mentioned sufficient type image-scanning device of the three-dimensional for being provided with depth camera
Point cloud P, Q of frame foot type image, and can further control the initial positional relationship of adjacent two frames foot type image.
102, the multiple characteristic point informations for obtaining described cloud P form feature point set Pi, the feature point set P includes described
The space coordinate of multiple characteristic points.
The pixel quantity of above-mentioned cloud P is in hundreds of thousands to millions of, if directly handling, calculation amount can be very big, compare consumption
When, therefore need to sample original point cloud P, that is, obtain the characteristic point information of point cloud;The method of sampling can be uniform sampling,
Method arrow sampling etc., to obtain features described above point set Pi, wherein subscript i indicates characteristic point i.
It is worth noting that features described above point set PiSpace coordinate including above-mentioned multiple characteristic points, this space coordinate institute
Coordinate system be using the position of above-mentioned depth camera as origin, camera institute towards direction be Z axis, two of the vertical plane of camera
Axial is X, Y-axis, the partial 3 d coordinate system of the camera of foundation;The space coordinate of features described above point is with this partial 3 d seat
X, Y, Z value based on mark system.
103, it is obtained on described cloud Q and corresponds to the feature point set PiFeature point set Qi, by transformation matrix by institute
State feature point set QiWith the feature point set PiIt is fitted, obtains fitting result, the fitting result includes error of fitting.
Wherein it is possible to by the methods of the distance of calculating point-to-point, point to the distance in face come the acquisition pair on above-mentioned cloud Q
Answer features described above point set PiFeature point set Qi, QiPartial 3 d coordinate including characteristic point;It then will be above-mentioned according to transformation matrix
Feature point set QiIn each point rotate a certain angle, then translate a certain distance and move closer to features described above point set PiPlace
Space coordinate, obtain that whether two feature point sets are overlapped as a result, and calculating above-mentioned error of fitting.
If 104, error of fitting is less than preset error threshold, changed according to above step to multiple image
In generation, obtains iteration result.
The error of fitting that step 103 obtains is compared with above-mentioned preset error threshold, if error of fitting
Less than error threshold, then explanation passes through above-mentioned two panels point cloud P, Q character pair point set P of above stepi、QiCorresponding points it is enough
It is close;Continue to be iterated multiple image repetition above step the iteration result for then obtaining multiple image.
105, according to the iteration result, the feature point set fusion figure of the multiple image is exported.
By step 104, above-mentioned iteration is carried out to the point cloud of multiple image, obtains and export the feature of above-mentioned multiple image
Point set fusion figure.
Further, as shown in figure 3, before executing above-mentioned steps 104, the present embodiment can with the following steps are included:
If 201, error of fitting is greater than the error threshold of setting, continue to obtain on described cloud Q corresponding to described
Feature point set PiFeature point set Qi, and update the transformation matrix;
202, pass through updated transformation matrix for the feature point set QiWith the feature point set PiIt is fitted, obtains
Fitting result, the fitting result include error of fitting.
For step 201, after obtaining error of fitting by step 103, by error of fitting and the error threshold of setting into
Row compares, if obtain error of fitting be greater than setting error threshold as a result, if by the distance that calculates point-to-point, point to face
The methods of distance continue to obtain corresponding features described above point set P on above-mentioned cloud QiFeature point set Qi, QiIncluding characteristic point
Partial 3 d coordinate;Then transformation matrix, including spin matrix and translation matrix are updated.
In step 202, after updating above-mentioned transformation matrix, by features described above point set QiBy new rotation angle peace move away from
From gradually close to features described above point set PiThe space coordinate at place obtains that whether two feature point sets are overlapped as a result, and calculating
New error of fitting out.
It is worth noting that above-mentioned transformation matrix includes initial value, that is, it include the initial value and translation matrix of spin matrix
Initial value, this initial value can determine by the fixed structure of hardware device, for example, can pass through above-mentioned three-dimensional sufficient type figure
The rocking bar uniform motion of depth camera is housed as the motor of the driving device of scanning means drives, continuous acquisition whithin a period of time
Depth image, in addition this parameter of the acquisition frame rate of depth camera, so that it may above-mentioned spin matrix and translation matrix be calculated
Initial value, specific process can be as follows: using above-mentioned three-dimensional sufficient type image-scanning device, depth camera is rotated around foot
270 degree, 10s is lasted, the acquisition frame rate of the depth camera is 30fps (frame is per second), can be calculated between adjacent two field pictures
Rotation angle be about:
That is, in the ideal case, adjacent two field pictures have rotated 0.9 degree, and translation distance 0m, the ideal situation refers to,
The motor rotation speed of above-mentioned scanning means is uniform, and each frame is identical relative to the angle that former frame rotates through;And by upper
An extension spring structure is arranged in the driving device for stating scanning means, makes seamless between reducer gear and motor gear, can make
Above-mentioned connecting rod rest operation is more steady, and the image of acquisition is more uniform;In scanning process the camera lens of depth camera to foot away from
Equal from always, the initial value of the transformation matrix of above-mentioned adjacent two frames foot-shape image available in this way, i.e., spin matrix is first
Initial value are as follows:
With the initial value of translation matrix are as follows: R0=0.
It is, of course, also possible to control the rotation speed of the motor of above-mentioned scanning means as needed and then control Image Acquisition
Time, come obtain different above-mentioned adjacent two frames foot-shape images transformation matrix initial value.
Further, as shown in figure 4, above-mentioned steps 103 may comprise steps of:
301, feature point set Q is up-sampled in described cloud Qi, so that QiCorresponding to PiThe sum of the distance between characteristic point it is small
In distance threshold, the distance threshold is calculated by least square method;
302, by feature point set QiBy transformation matrix rotation translation after with the feature point set PiIt is fitted, obtains described
Error of fitting.
For step 301, corresponding P is determined on cloud QiFeature point set QiWhen, using point-to-point apart from this side
Method, i.e. air line distance between corresponding points;Above-mentioned distance threshold is that is, to make P by least squarei、QiCorresponding points
The coordinate value of (i.e. i-th point) subtracts each other rear quadratic sum minimum to be calculated, and formula is as follows:
Wherein, subscript i indicates that ith feature point, k represent the number of iterations.
Therefore step 301 samples obtained feature point set Qi, QiEach characteristic point to PiStraight line between corresponding characteristic point
Sum of the distance is less than Pi、QiThe coordinate value of corresponding points subtracts each other the minimum value of rear quadratic sum.
For step 302, according to transformation matrix by the feature point set Q of above-mentioned determinationiIn each point rotate certain angle
Degree, then translate a certain distance and move closer to features described above point set PiWhether the space coordinate at place obtains two feature point sets
Be overlapped as a result, and calculating above-mentioned error of fitting.
Further, the update transformation matrix in above-mentioned steps 201 may include:
Transformation matrix is calculated and updated according to above-mentioned distance threshold, i.e., by least square, that is, makes above-mentioned Pi、
QiThe coordinate value of (after transformation) corresponding points subtracts each other rear quadratic sum minimum to obtain new transformation matrix, and formula is as follows:
Further, so that error of fitting is less than preset error threshold described in above-mentioned steps 104, can wrap
It includes:
By above-mentioned transformation matrix by feature point set QiGradually rotation and translation, so that features described above point set QiTo feature point set
PiMean longitudinal error be less than above-mentioned preset error threshold.
It should be noted that the step for transformation matrix have been subjected to update;PiTo QiMean longitudinal error by Pi、Qi
The coordinate value of corresponding points subtracts each other rear square and then summation takes average error to obtain again, and formula is as follows:
In the embodiment of the present invention, point cloud P, Q of adjacent two frames foot type image are obtained respectively;Obtain the multiple of described cloud P
Characteristic point information forms feature point set Pi, the feature point set PiSpace coordinate including the multiple characteristic point;In described cloud
It is obtained on Q and corresponds to the feature point set PiFeature point set Qi, by transformation matrix by the feature point set QiWith the feature
Point set PiIt is fitted, obtains fitting result, the fitting result includes error of fitting;It is preset if error of fitting is less than
Error threshold, then multiple image is iterated according to above step, obtains iteration result;It is defeated according to the iteration result
The feature point set of the multiple image merges figure out.The initial transformation square obtained due to the point cloud by adjacent two frames foot type image
Battle array is smaller, the number of point-cloud fitting iteration is reduced, so as to improve the speed of image mosaic.
Fig. 6 is referred to, Fig. 6 is a kind of structural representation of three-dimensional sufficient type image processing apparatus provided in an embodiment of the present invention
Figure, as shown in fig. 6, described device includes:
First obtains module 501, for obtaining point cloud P, Q of adjacent two frames foot type image respectively;
Second obtains module 502, and multiple characteristic point informations for obtaining described cloud P form feature point set Pi, described
Feature point set PiSpace coordinate including the multiple characteristic point;
Fitting module 503 corresponds to the feature point set P for obtaining on described cloud QiFeature point set Qi, pass through
Transformation matrix is by the feature point set QiWith the feature point set PiIt is fitted, obtains fitting result, the fitting result includes
Error of fitting;
Judgment module 504, if being less than preset error threshold for error of fitting, according to above step to more
Frame image is iterated, and obtains iteration result;
Output module 505, for exporting the feature point set fusion figure of the multiple image according to the iteration result.
A kind of three-dimensional sufficient type image processing apparatus provided in an embodiment of the present invention can be realized in the embodiment of the method for Fig. 1
Each embodiment and corresponding beneficial effect, to avoid repeating, which is not described herein again.
It is the structural schematic diagram of a kind of electronic equipment provided in an embodiment of the present invention referring to Fig. 5, Fig. 5, as shown in figure 4, packet
It includes: memory 402, processor 401 and being stored in the calculating that can be run on the memory 402 and on the processor 401
Machine program, in which:
The computer program that processor 401 is used to that memory 402 to be called to store executes following steps:
Point cloud P, Q of adjacent two frames foot type image are obtained respectively;
The multiple characteristic point informations for obtaining described cloud P form feature point set Pi, the feature point set PiIncluding the multiple
The space coordinate of characteristic point;
It is obtained on described cloud Q and corresponds to the feature point set PiFeature point set Qi, by transformation matrix by the spy
Levy point set QiWith the feature point set PiIt is fitted, obtains fitting result, the fitting result includes error of fitting;
If error of fitting is less than preset error threshold, multiple image is iterated according to above step,
Obtain iteration result;
According to the iteration result, the feature point set fusion figure of the multiple image is exported.
Optionally, if being less than preset error threshold in the error of fitting, according to above step to multiframe
Before the step for image is iterated, and obtains iteration result, the method also includes steps:
If error of fitting is greater than the error threshold of setting, continue to obtain on described cloud Q corresponding to the feature
Point set PiFeature point set Qi, and update the transformation matrix;
By updated transformation matrix by the feature point set QiWith the feature point set PiIt is fitted, is fitted
As a result, the fitting result includes error of fitting.
Optionally, the transformation matrix includes initial value, i.e., the initial value including spin matrix and translation matrix is initial
Value.
Optionally, the initial value of the transformation matrix, the i.e. initial value of spin matrix and the initial value of translation matrix pass through
The fixed structure of hardware device determines.
Processor 401 executes described obtain on described cloud Q and corresponds to the feature point set PiFeature point set Qi, lead to
Transformation matrix is crossed by the feature point set QiWith the feature point set PiIt is fitted, obtains fitting result, the fitting result packet
Include error of fitting, comprising:
Feature point set Q is up-sampled in described cloud Qi, so that QiCorresponding to PiThe sum of the distance between characteristic point be less than away from
From threshold value, the distance threshold is calculated by least square method;
By feature point set QiBy transformation matrix rotation translation after with the feature point set PiIt is fitted, obtains the fitting
Error.
The update transformation matrix that processor 401 executes, comprising:
It is calculated according to the distance threshold and updates the transformation matrix.
The described of the execution of processor 401 makes error of fitting be less than preset error threshold, comprising:
By the transformation matrix by feature point set QiGradually rotation and translation, so that the feature point set QiTo the feature
Point set PiMean longitudinal error be less than the error threshold.
Above-mentioned processor 401 can be in some embodiments central processing unit (Central Processing Unit,
CPU), controller, microcontroller, microprocessor or other data processing chips.
It should be noted that since above-mentioned processor 401 executes the computer program that meter is stored in above-mentioned memory 402
When the step of can realizing above-mentioned three-dimensional foot-shape image processing method, therefore all realities of above-mentioned three-dimensional foot-shape image processing method
It applies example and is suitable for above-mentioned electronic equipment, and can reach the same or similar beneficial effect.
In addition, specific embodiments of the present invention additionally provide a kind of computer readable storage medium 402, it is computer-readable to deposit
Storage media 402 is stored with computer program, which realizes when being executed by processor at above-mentioned three-dimensional foot-shape image
The step of reason method.
That is, in a specific embodiment of the present invention, the computer program of computer readable storage medium is executed by processor
The step of Shi Shixian above-mentioned three-dimensional foot-shape image processing method, it can be improved the speed of image procossing.
Illustratively, the computer program of computer readable storage medium includes computer program code, the computer
Program code can be source code form, object identification code form, executable file or certain intermediate forms etc..The computer can
Reading medium may include: any entity or device, recording medium, USB flash disk, mobile hard that can carry the computer program code
Disk, magnetic disk, CD, computer storage, read-only memory (ROM, Read-Only Memory), random access memory
(RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium etc..
It should be noted that the computer program due to computer readable storage medium realized when being executed by processor it is above-mentioned
Three-dimensional foot-shape image processing method the step of, therefore all embodiments of above-mentioned three-dimensional foot-shape image processing method are suitable for
The computer readable storage medium, and can reach the same or similar beneficial effect.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the program can be stored in a computer-readable storage medium
In, the program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.
It should be noted that for the various method embodiments described above, for simple description, therefore, it is stated as a series of
Combination of actions, but those skilled in the art should understand that, the application is not limited by the described action sequence because
According to the application, some steps may be performed in other sequences or simultaneously.Secondly, those skilled in the art should also know
It knows, embodiment described in this description belongs to alternative embodiment, related actions and modules not necessarily the application
It is necessary.
In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, there is no the portion being described in detail in some embodiment
Point, reference can be made to the related descriptions of other embodiments.
In several embodiments provided herein, it should be understood that disclosed device, it can be by another way
It realizes.For example, the apparatus embodiments described above are merely exemplary, such as the division of the unit, it is only a kind of
Logical function partition, there may be another division manner in actual implementation, such as multiple units or components can combine or can
To be integrated into another system, or some features can be ignored or not executed.Another point, shown or discussed is mutual
Coupling, direct-coupling or communication connection can be through some interfaces, the indirect coupling or communication connection of device or unit,
It can be electrical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, each functional unit in each embodiment of the application can integrate in one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also be realized in the form of software program module.
If the integrated unit is realized in the form of software program module and sells or use as independent product
When, it can store in a computer-readable access to memory.Based on this understanding, the technical solution of the application substantially or
Person says that all or part of the part that contributes to existing technology or the technical solution can body in the form of software products
Reveal and, which is stored in a memory, including some instructions are used so that a computer equipment
(can be personal computer, server or network equipment etc.) executes all or part of each embodiment the method for the application
Step.And memory above-mentioned includes: USB flash disk, read-only memory (ROM, Read-Only Memory), random access memory
The various media that can store program code such as (RAM, Random Access Memory), mobile hard disk, magnetic or disk.
Those of ordinary skill in the art will appreciate that all or part of the steps in the various methods of above-described embodiment is can
It is completed with instructing relevant hardware by program, which can store in a computer-readable memory, memory
May include: flash disk, read-only memory (English: Read-Only Memory, referred to as: ROM), random access device (English:
Random Access Memory, referred to as: RAM), disk or CD etc..
The above disclosure is only the preferred embodiments of the present invention, cannot limit the right model of the present invention with this certainly
It encloses, therefore equivalent changes made in accordance with the claims of the present invention, is still within the scope of the present invention.
Claims (10)
1. a kind of three-dimensional sufficient type image processing method characterized by comprising
Point cloud P, Q of adjacent two frames foot type image are obtained respectively;
The multiple characteristic point informations for obtaining described cloud P form feature point set Pi, the feature point set PiIncluding the multiple feature
The space coordinate of point;
It is obtained on described cloud Q and corresponds to the feature point set PiFeature point set Qi, by transformation matrix by the characteristic point
Collect QiWith the feature point set PiIt is fitted, obtains fitting result, the fitting result includes error of fitting;
If error of fitting is less than preset error threshold, multiple image is iterated according to above step, is obtained
Iteration result;
According to the iteration result, the feature point set fusion figure of the multiple image is exported.
2. the method as described in claim 1, which is characterized in that if being less than preset error threshold in the error of fitting
Before the step for being worth, being then iterated to multiple image according to above step, obtain iteration result, further comprise the steps of:
If error of fitting is greater than the error threshold of setting, continue to obtain on described cloud Q corresponding to the feature point set Pi
Feature point set Qi, and update the transformation matrix;
By updated transformation matrix by the feature point set QiWith the feature point set PiIt is fitted, obtains fitting result,
The fitting result includes error of fitting.
3. method according to claim 2, which is characterized in that the transformation matrix includes spin matrix and translation matrix, institute
Stating spin matrix includes rotation initial value, and the translation matrix includes translation initial value.
4. method as claimed in claim 3, which is characterized in that according to the angle difference of the adjacent two frames foot type picture point cloud
Obtaining the rotation initial value is, obtains the translation initial value according to the distance difference of the adjacent two frames foot type picture point cloud
For.
5. method as claimed in claim 4, which is characterized in that described obtain on described cloud Q corresponds to the characteristic point
Collect PiFeature point set Qi, by transformation matrix by the feature point set QiWith the feature point set PiIt is fitted, is fitted
As a result, the fitting result includes error of fitting, comprising:
Feature point set Q is up-sampled in described cloud Qi, so that QiCorresponding to PiThe sum of the distance between characteristic point be less than apart from threshold
Value, the distance threshold are calculated by least square method;
By feature point set QiBy transformation matrix rotation translation after with the feature point set PiIt is fitted, obtains the error of fitting.
6. method as claimed in claim 5, which is characterized in that described to update the transformation matrix, comprising:
The transformation matrix is obtained according to the distance threshold.
7. method as claimed in claim 6, which is characterized in that if error of fitting is less than preset error threshold,
Multiple image is iterated according to above step, obtains iteration result, comprising:
By the transformation matrix by feature point set QiRotation and translation obtains the feature point set QiTo the feature point set Pi's
Mean longitudinal error;
If the mean longitudinal error is less than the preset error threshold, continue to be iterated next frame image;
If the mean longitudinal error is greater than the preset error threshold, continue to correspond in described cloud Q up-sampling
In the feature point set PiFeature point set Qi。
8. a kind of device of three-dimensional sufficient type image processing method characterized by comprising
First obtains module, for obtaining point cloud P, Q of adjacent two frames foot type image respectively;
Second obtains module, and multiple characteristic point informations for obtaining described cloud P form feature point set Pi, the feature point set
PiSpace coordinate including the multiple characteristic point;
Fitting module corresponds to the feature point set P for obtaining on described cloud QiFeature point set Qi, by converting square
Battle array is by the feature point set QiWith the feature point set PiIt is fitted, obtains fitting result, the fitting result includes that fitting misses
Difference;
Judgment module, if being less than preset error threshold for error of fitting, according to above step to multiple image
It is iterated, obtains iteration result;
Output module, for exporting the feature point set fusion figure of the multiple image according to the iteration result.
9. a kind of electronic equipment characterized by comprising memory, processor and be stored on the memory and can be in institute
The computer program run on processor is stated, the processor is realized when executing the computer program as in claim 1 to 7
Step in described in any item three-dimensional sufficient type image processing methods.
10. a kind of computer readable storage medium, which is characterized in that be stored with computer on the computer readable storage medium
Program realizes the three-dimensional sufficient type image as described in any one of claims 1 to 7 when the computer program is executed by processor
Step in processing method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811570138.XA CN109785278B (en) | 2018-12-21 | Three-dimensional foot-type image processing method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811570138.XA CN109785278B (en) | 2018-12-21 | Three-dimensional foot-type image processing method and device, electronic equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109785278A true CN109785278A (en) | 2019-05-21 |
CN109785278B CN109785278B (en) | 2024-05-31 |
Family
ID=
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111739071A (en) * | 2020-06-15 | 2020-10-02 | 武汉尺子科技有限公司 | Rapid iterative registration method, medium, terminal and device based on initial value |
CN113538552A (en) * | 2020-02-17 | 2021-10-22 | 天目爱视(北京)科技有限公司 | 3D information synthesis image matching method based on image sorting |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130322698A1 (en) * | 2010-10-01 | 2013-12-05 | Saab Ab | Method and an apparatus for image-based navigation |
CN104574401A (en) * | 2015-01-09 | 2015-04-29 | 北京环境特性研究所 | Image registration method based on parallel line matching |
CN106157367A (en) * | 2015-03-23 | 2016-11-23 | 联想(北京)有限公司 | Method for reconstructing three-dimensional scene and equipment |
CN106251399A (en) * | 2016-08-30 | 2016-12-21 | 广州市绯影信息科技有限公司 | A kind of outdoor scene three-dimensional rebuilding method based on lsd slam |
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130322698A1 (en) * | 2010-10-01 | 2013-12-05 | Saab Ab | Method and an apparatus for image-based navigation |
CN104574401A (en) * | 2015-01-09 | 2015-04-29 | 北京环境特性研究所 | Image registration method based on parallel line matching |
CN106157367A (en) * | 2015-03-23 | 2016-11-23 | 联想(北京)有限公司 | Method for reconstructing three-dimensional scene and equipment |
CN106251399A (en) * | 2016-08-30 | 2016-12-21 | 广州市绯影信息科技有限公司 | A kind of outdoor scene three-dimensional rebuilding method based on lsd slam |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113538552A (en) * | 2020-02-17 | 2021-10-22 | 天目爱视(北京)科技有限公司 | 3D information synthesis image matching method based on image sorting |
CN113538552B (en) * | 2020-02-17 | 2024-03-22 | 天目爱视(北京)科技有限公司 | 3D information synthetic image matching method based on image sorting |
CN111739071A (en) * | 2020-06-15 | 2020-10-02 | 武汉尺子科技有限公司 | Rapid iterative registration method, medium, terminal and device based on initial value |
CN111739071B (en) * | 2020-06-15 | 2023-09-05 | 武汉尺子科技有限公司 | Initial value-based rapid iterative registration method, medium, terminal and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108269279B (en) | Three-dimensional reconstruction method and device based on monocular 3 D scanning system | |
US9418474B2 (en) | Three-dimensional model refinement | |
CN104574501B (en) | A kind of high-quality texture mapping method for complex three-dimensional scene | |
JP4679033B2 (en) | System and method for median fusion of depth maps | |
CN111784821B (en) | Three-dimensional model generation method and device, computer equipment and storage medium | |
WO2016081722A1 (en) | Systems and methods for 3d capture of objects using multiple range cameras and multiple rgb cameras | |
JP4889182B2 (en) | Depth map generation by hypothesis mixing in Bayesian framework | |
CN109191509A (en) | A kind of virtual binocular three-dimensional reconstruction method based on structure light | |
CN108734776A (en) | A kind of three-dimensional facial reconstruction method and equipment based on speckle | |
CN106464851A (en) | Depth estimation using multi-view stereo and a calibrated projector | |
CN104255026B (en) | Image processing equipment, information processing equipment and image processing method | |
CN110349257B (en) | Phase pseudo mapping-based binocular measurement missing point cloud interpolation method | |
CN110378967B (en) | Virtual target calibration method combining grating projection and stereoscopic vision | |
CN108460824B (en) | Method, device and system for determining stereoscopic multimedia information | |
US10909704B2 (en) | Apparatus and a method for generating data representing a pixel beam | |
CN109785278A (en) | A kind of three-dimensional sufficient type image processing method, device, electronic equipment and storage medium | |
CN109218706B (en) | Method for generating stereoscopic vision image from single image | |
CN116012449A (en) | Image rendering method and device based on depth information | |
CN109712230A (en) | Threedimensional model compensation process, device, storage medium and processor | |
CN110149508A (en) | A kind of array of figure generation and complementing method based on one-dimensional integrated imaging system | |
CN112967370B (en) | Three-dimensional light field reconstruction method and device and storage medium | |
CN109816765A (en) | Texture towards dynamic scene determines method, apparatus, equipment and medium in real time | |
CN109785278B (en) | Three-dimensional foot-type image processing method and device, electronic equipment and storage medium | |
CN111754615B (en) | Real-time reconstruction method and device for high-quality textures | |
CN111161400B (en) | Glasses matching design equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant |