CN109829927A - A kind of electronic glasses and high-altitude scene image method for reconstructing - Google Patents
A kind of electronic glasses and high-altitude scene image method for reconstructing Download PDFInfo
- Publication number
- CN109829927A CN109829927A CN201910095865.3A CN201910095865A CN109829927A CN 109829927 A CN109829927 A CN 109829927A CN 201910095865 A CN201910095865 A CN 201910095865A CN 109829927 A CN109829927 A CN 109829927A
- Authority
- CN
- China
- Prior art keywords
- image
- midpoint
- straight line
- glasses
- altitude
- 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
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a kind of electronic glasses and high-altitude scene image method for reconstructing, are related to the technical field that computer vision is combined with sensor.The present invention is acquired reality scene and head pose by binocular camera and six axis attitude transducers, and it is transmitted to master controller, master controller converts algorithm by OpenCV_BM algorithm and Hough straight line and analyzes image data, it is partitioned into far and near image, close shot and material particular are retained, distant view is handled using the methods of Gaussian Blur, and finally obtained image is shown on liquid crystal display, and whole process can reach good real-time.When wearer is in eminence and wears the electronic glasses, it is seen that distant view image be blurred picture, thus effectively alleviate wearer acrophobia shape；And the electronic glasses have constructively used modularized design, and do not need expensive optical mirror slip, to reduce maintenance cost and manufacturing cost.
Description
Technical field
The present invention relates to the technical field that computer vision is combined with sensor, more particularly to a kind of electronic glasses and
High-altitude scene image method for reconstructing.
Background technique
Acrophobia also known as fears high disease.Data is shown according to investigations, has 91% people acrophobia shape occurred in people from modern metropolitan cities.
Wherein 10% belongs to clinical probably height.It is learnt by inquiry data, the treatment for acrophobe is largely all based at heart
Treatment, more commonly used method are that exposure method is also known as flooding therapy, and this method is to encourage patient receiving treatment straight
Contact is led to the scene of terrified anxiety, adheres to a kind of quick behaviour therapy method that strain sensation disappears.This treatment method
Treatment time it is long while can give patient receiving treatment very big pressure at heart, so being ready to receive the people for the treatment of still
It is relatively fewer.In addition to this, it additionally provides a kind of optics to subtract away from glasses, but the cost of this kind of glasses is relatively high, while effect
It is not significant.Therefore, there are no be absorbed in solution or alleviate the various fields that encounter in real life of acrophobe on the market
Scape and the product of fear generated.
Summary of the invention
In order to make up the defect of the prior art, the present invention provides a kind of electronic glasses and high-altitude scene image reconstruction sides
Method.
To achieve the above object, the present invention provides following schemes:
A kind of electronic glasses, including glasses box and binocular camera；The glasses box is embedded with the liquid set gradually
Crystal display screen, master control borad and mobile power source；Master controller and six axis attitude transducers are integrated on the master control borad；It is described double
Mesh camera is mounted on the lateral surface of the glasses box, the binocular camera, the six axis attitude transducer, the shifting
Dynamic power supply is electrically connected with the master controller.
Optionally, it is embedded with sighting distance at the wearing port of the glasses box and adjusts eyeglass, and the sighting distance adjusts mirror
Piece is located at the front of the liquid crystal display；The lateral surface of the glasses box is additionally provided with sighting distance and adjusts sliding shoe.
Optionally, the wearing port of the glasses box is arcuate structure.
Optionally, the glasses box is assembled using clamping plate mode, and the material of the glasses box is PLA material.
Optionally, the master controller reaches the Intel Cherry Trail Z8350 tetra- of 1.44GHz using base frequency
Core processor；The master controller is integrated with HDMI video output port and three USB ports；The master controller carries open source
Ubuntu system and support OpenCV_BM program.
Optionally, the binocular camera is made of two camera modules, and the binocular camera is defeated using single USB
Out, and Windows system and linux system are compatible with；The video output formats of the binocular camera are yuv format.
Optionally, the six axis attitude transducer is gyroscope accelerometer.
Optionally, the model MPU6050 of the gyroscope accelerometer.
A kind of high-altitude scene image method for reconstructing applied to electronic glasses, the high-altitude scene image method for reconstructing packet
It includes:
The high-altitude scene image of binocular camera acquisition and the head pose information of six axis attitude transducers acquisition are obtained,
And determine the head rotation angle in the head pose information；
Judge that the head rotation angle whether in set angle section, obtains the first judging result；
If first judging result indicates the head rotation angle not in set angle section, then by the height
Empty scene image is shown on liquid crystal display；
If first judging result indicates the head rotation angle in set angle section, then by the high-altitude
Scene image is converted into bianry image, and carries out edge detection to each region in the bianry image by Canny algorithm,
Obtain the boundary curve of all areas；
All boundary curves are detected using Hough straight line transformation algorithm, straight line is judged whether there is, obtains
Second judging result；
If second judging result indicates that there are straight lines in the boundary curve, calculate the midpoint of all straight lines, so
Segmentation contour is determined according to the distributing position at the midpoint of all straight lines afterwards, and according to the segmentation contour to the high-altitude scene figure
As being split, close shot image and distant view image are obtained；
If second judging result indicates that straight line is not present in the boundary curve, OpenCV_BM algorithm pair is used
The high-altitude scene image is handled to obtain depth map, is then carried out according to erosion algorithm and expansion algorithm to the depth map
Processing obtains the grayscale image of different depth, and according to the gray threshold of setting, carries out binaryzation division to the grayscale image, obtain
To close shot image and distant view image；The distant view image is gray value corresponding high-altitude scene image region when being 0, the close shot
Image is gray value corresponding high-altitude scene image region when being 1；Wherein, it will be above the gray value of the pixel of the gray threshold
Become 0, the gray value that will be less than the pixel of the gray threshold becomes 1.
Fuzzy Processing is carried out to the distant view image；
By the distant view image synthesis after the close shot image and Fuzzy Processing, and composograph is shown in liquid crystal display
On.
Optionally, the distributing position at the midpoint according to all straight lines determines segmentation contour, specifically includes:
The bianry image is divided into three parts, respectively left part, middle part and right part according to X direction；If distribution
Quantity at the midpoint of left part is greater than first threshold and filters out the horizontal seat at midpoint then according to the coordinate value at the midpoint of all straight lines
Maximum value is marked, and the corresponding midpoint coordinates of abscissa maximum value is determined as to divide the midpoint coordinates of straight line；If being distributed in right part
The quantity at midpoint be greater than the first threshold and then according to the coordinate value at the midpoint of all straight lines filter out the abscissa at midpoint
Minimum value, and the corresponding midpoint coordinates of abscissa minimum value is determined as to divide the midpoint coordinates of straight line；
The bianry image is divided into three parts, respectively left part, middle part and right part according to horizontal axis method；If distribution
It is greater than second threshold in the quantity at middle part and the midpoint of lower part and filters out midpoint then according to the coordinate value at the midpoint of all straight lines
Ordinate minimum value, and by the corresponding midpoint coordinates of ordinate minimum value be determined as divide straight line midpoint coordinates；
According to the midpoint coordinates of the segmentation straight line, segmentation straight line is calculated；
The binary image is handled according to findcontours () function, obtains the edge of all enclosed regions
Curve；
Straight line will be divided to connect with the boundary curve of the segmentation immediate enclosed region of straight line endpoint, determine dividing wheel
It is wide.
The specific embodiment provided according to the present invention, the invention discloses following technical effects:
The present invention provides a kind of electronic glasses and high-altitude scene image method for reconstructing, pass through binocular camera and six axis
Attitude transducer is acquired reality scene and head pose, these data are then transmitted to master controller, and master controller is first
Head pose is analyzed, if it is looking squarely or looking up posture, image is not handled；If overlooking or shape obliquely
State then analyzes image data by OpenCV_BM algorithm and Hough transformation algorithm, far and near image is partitioned into, to close
Scape image and material particular are retained, and distant view image is handled using the methods of Gaussian Blur, finally obtained shadow
As being shown on liquid crystal display, and entire treatment process has reached good real-time.When wearer is in eminence and pendant
When wearing the electronic glasses, it is seen that distant view image be blurred picture, thus effectively alleviate wearer acrophobia shape；And the electricity
Sub- glasses have constructively used modularized design, and do not need expensive optical mirror slip, to reduce maintenance cost and system
Cause this.
Detailed description of the invention
It in order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, below will be to institute in embodiment
Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the invention
Example, for those of ordinary skill in the art, without any creative labor, can also be according to these attached drawings
Obtain other attached drawings.
Fig. 1 is the structural schematic diagram of glasses of embodiment of the present invention box；
Fig. 2 is the front view of glasses of embodiment of the present invention box；
Fig. 3 is the rearview of glasses of embodiment of the present invention box；
Fig. 4 is the top view of glasses of embodiment of the present invention box；
Fig. 5 is the flow diagram of high-altitude of embodiment of the present invention scene image method for reconstructing.
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 only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
The object of the present invention is to provide a kind of electronic glasses and high-altitude scene image method for reconstructing, pass through computer vision skill
Art will cause acrophobe to generate fear elevation information in picture weakens, but retains some important in picture
Scenery details allows wearer that can understand the position where oneself, such as on high mountain, on skyscraper, on roof balcony etc., from
And slow down acrophobe's feared state of mind that scene generates in real life.
And the electronic glasses have constructively used modularized design, and do not need expensive optical mirror slip, to reduce
Maintenance cost and manufacturing cost.
In order to make the foregoing objectives, features and advantages of the present invention clearer and more comprehensible, with reference to the accompanying drawing and specific real
Applying mode, the present invention is described in further detail.
This system is related to hardware circuit, core algorithm and structural model.Hardware circuit includes: master cpu, binocular camera shooting
Head, six axis attitude transducers；Structural model is the wear-type box for being similar to VR, can load all hardware circuit.
Electronic glasses provided in this embodiment include glasses box and binocular camera；Glasses box, which is embedded with, to be set gradually
Liquid crystal display, master control borad and mobile power source；Master controller and six axis attitude transducers are integrated on master control borad；Binocular is taken the photograph
It is mounted on as head on the lateral surface of glasses box, binocular camera, six axis attitude transducers, mobile power source are electric with master controller
Connection.
It is embedded with sighting distance at the wearing port of glasses box and adjusts eyeglass, and sighting distance adjusts eyeglass and is located at liquid crystal display
Front；Sighting distance is adjusted eyeglass to be arranged in glasses box, glasses box is made to can adapt to all ages and classes, different eyesight shapes
The crowd of condition.
The lateral surface of glasses box is additionally provided with sighting distance and adjusts sliding shoe, switch key and restart key.Glasses box
Wearing port is arcuate structure.
Glasses box is assembled using clamping plate mode.The material of glasses box is PLA material.
Preferably, the six axis attitude transducers that the present embodiment uses is gyroscope accelerometers.
It describes in detail below to the various components in electronic glasses
Master controller
The present embodiment uses the mini plate of the higher LattePanda of cost performance.Specifically: master controller uses Intel
Tetra- core processor of Cherry Trail Z8350, base frequency reach 1.44GHz, and are integrated with three USB ports and HDMI view
Frequency exports, and can carry the Ubuntu system of open source and support Open_CV program, be very suitable to the exploitation of this system.
Design feature: (only 88*70mm*16mm) small in size, light-weight (55 grams), integrated level is high.
Electric characteristic: 5V/2A power supply, low in energy consumption (SDP of whole system chip is 2 watts, and thermal design power should be lower than 4
Watt).
Binocular camera
Electronic glasses provided in this embodiment need to carry out the extraneous world three-dimensional perception, it is contemplated that cost and image are real
The factors such as when property, use a common binocular vision camera on the market.The binocular camera 2 is by two camera module groups
At, it is exported using single USB, reduces the occupancy to USB port, compatible each large platform of Windows, Linux, video output formats
Reduce the occupancy to CPU without decoding for yuv format.
Compared to the other methods such as laser sensor ranging and monocular ranging, there is high-efficient, precision to close for binocular ranging
Suitable, the advantages that system structure is simple, at low cost.
Gyroscope accelerometer
The gyroscope accelerometer model MPU6050 that the present embodiment uses, identifies human body head posture, this gyro
Instrument accelerometer uses advanced digital filtering technique, and measurement noise can be effectively reduced, and improves measurement accuracy.
Design feature: integrated level is high, small in size, light-weight.
Electric characteristic: low in energy consumption, operating current is smaller (less than 10 milliamperes).
Liquid crystal display
The present embodiment uses 5 cun of LCD displays of micro- snow, and resolution ratio 800*480 is compatible stronger.
Mobile power source
The problem of in view of continuation of the journey and carrying, the present embodiment has selected mobile power source common to be on the market powered,
Mobile power source has many advantages, such as that capacity is big, volume is suitable, easy to carry, can meet the power supply of this equipment, can also meet mobile phone
Deng the charge requirement of other smart machines.
Glasses box
As shown in Figs 1-4, glasses box provided in this embodiment is with reference to VR device housings on the market, by multiple
Test, upgrading improve, and have redesigned the glasses box of a suitable human eye direct viewing.Spectacle case is set using clamping plate mode
Meter is moved easily the installation and removal of power supply and other equipment, and the material of glasses box uses PLA material, and is wearing
Using personalized cambered design at port, can more comfortably wear.
The glasses box 1 includes upper top panel, bottom panel, front panel, rear panel 1 and Left-Hand Panel 2 and right side
Plate.Handle is equipped on Left-Hand Panel 2 and right panel.
In the present embodiment, sighting distance adjusts sliding shoe 3, switch key 4 and restarts the upper top that glasses box is arranged in key 5
On panel, there are two camera mounting holes 6, binocular camera to be mounted on camera mounting hole for processing on the front panel of glasses box
It is fixed in 6.It wears port to be mounted on the rear panel 1 of glasses box, and is also processed on rear panel 1 there are two eyeglass mounting hole 7,
Sighting distance adjusting eyeglass is mounted in eyeglass mounting hole 7 fixed.
Three pieces of clamping plates, respectively display screen clamping plate 8, processor clamping plate 9 and power supply clamping plate have been embedded in glasses box 1
10, when installation only need liquid crystal display, master control borad and mobile power source insertion respectively display screen clamping plate 8, processor clamping plate 9 and
In power supply clamping plate 10.
Fig. 5 is the flow diagram of high-altitude of embodiment of the present invention scene image method for reconstructing, as shown in figure 5, the present embodiment
A kind of high-altitude scene image method for reconstructing is provided, following steps are specifically included.
Step 101: obtaining the high-altitude scene image of binocular camera acquisition and the head of six axis attitude transducers acquisition
Posture information, and determine the head rotation angle in the head pose information.
Step 102: judging that the head rotation angle whether in set angle section, obtains the first judging result；If
It is that first judging result indicates that the head rotation angle not in set angle section, thens follow the steps 103；If institute
Stating the first judging result indicates that the head rotation angle in set angle section, thens follow the steps 104.
Step 103: the high-altitude scene image is shown on liquid crystal display.
Step 104: the high-altitude scene image being converted into bianry image, and by Canny algorithm to the binary map
Each region as in carries out edge detection, obtains the boundary curve of all areas.
Step 105: all boundary curves being detected using Hough straight line transformation algorithm, are judged whether there is straight
Line obtains the second judging result.If second judging result indicates that there are straight lines in the boundary curve, then follow the steps
106；If second judging result indicates that straight line is not present in the boundary curve, 107 are thened follow the steps.
Step 106: calculating the midpoint of all straight lines, dividing wheel is then determined according to the distributing position at the midpoint of all straight lines
Exterior feature, and the high-altitude scene image is split according to the segmentation contour, obtain close shot image and distant view image.
Step 107: the high-altitude scene image being handled to obtain depth map using OpenCV_BM algorithm, then root
The depth map is handled according to erosion algorithm and expansion algorithm, obtains the grayscale image of different depth, and according to the ash of setting
Threshold value is spent, binaryzation division is carried out to the grayscale image, obtains close shot image and distant view image；The distant view image is gray value
Corresponding high-altitude scene image region when being 0, the close shot image is gray value corresponding high-altitude scene image region when being 1；
Wherein, the gray value that will be above the pixel of the gray threshold becomes 0, and the gray value that will be less than the pixel of the gray threshold becomes
At 1.
Step 108: Fuzzy Processing is carried out to the distant view image.
Step 109: the distant view image after the close shot image and Fuzzy Processing being synthesized, and composograph is shown in liquid
On crystal display screen.
After high-altitude scene image is converted to grayscale image, gaussian filtering is realized by convolution of a two-dimensional Gaussian kernel,
Then amplitude and the direction that gradient is calculated with the finite difference of single order local derviation carry out non-maxima suppression to the amplitude of gradient,
That is finding pixel local maximum, the corresponding gray value of non-maximum value is set to 0, completes non-maxima suppression
Afterwards, a bianry image can be obtained.The gray value of the point of non-edge is 0, may be the local gray level maximum of the point at edge
The gray value of point setting is 128.
The bianry image further comprises much vacation edge (closures in non-bianry image as caused by noise and other reasons
Region contour edge).The present embodiment reduces false amount of edge using the dual threshold method of Canny algorithm, specially selects two
A threshold value, respectively high threshold and Low threshold obtain a region contour according to high threshold, and such a region contour contains very
Few false edge, but since high threshold is higher, the region contour of generation may not be closed, it is therefore desirable to Low threshold goes to solve area
The inc problem of domain profile.
The segmentation contour of bianry image is defined below, that just needs to search out the watershed of far point and near point.Similar to ground
Horizontal line, can be by sky and surface separation with horizon.The watershed of far point and near point can not be directly determined at present.If to two
Value image directly finds enclosed region profile, will be interfered by the enclosed region profile in far point, so before this, it be first
Substantially determine the position in the watershed of far point near point.
By actual observation and use, in the picture, the edge of far and near image is different to the far and near reflection in the visual field
's.The edge of distant view image be mostly it is desultory, it is incomplete.And the edge of close shot image be all it is continuous complete,
And most edge lines.So being next then to find straight line present in bianry image according to this feature.The present invention is implemented
Example carries out straight-line detection to the edge in bianry image using Hough straight line transformation algorithm.
Wherein, segmentation contour is determined according to the distributing position at the midpoint of all straight lines, specifically included: is distributed by midpoint
Probability is divided into three kinds: the first, close shot is below the scene image of high-altitude；Second, close shot is on the left side of high-altitude scene image；
The third, close shot is on the right of high-altitude scene image.The foundation of classification is the institute in three regions the case where calculating midpoint distribution
The midpoint for including is more, then it is assumed that close shot is just in this region.
Concrete thought are as follows: bianry image is divided into three parts, respectively left part, middle part and the right side according to X direction
Portion.If midpoint is largely distributed in left part, belong to above-mentioned second situation, if midpoint is largely distributed in right part, belongs to
The third above-mentioned situation.Bianry image is divided into three parts according to y direction, respectively top, middle part and lower part, if in
Point is largely distributed in middle and lower part, then belongs to the first situation.If it is the first situation, subsequent processing is then that screening is all straight
The corresponding midpoint coordinates of ordinate minimum value in line midpoint.Second situation is then that x abscissa is most in all straight line midpoints of screening
It is worth corresponding midpoint coordinates greatly.The third situation is then that the corresponding midpoint of x abscissa minimum value is sat in all straight line midpoints of screening
Mark.Then segmentation straight line is calculated according to this midpoint coordinates.By this thinking, it can be recognized accurately and belong to the straight of close shot edge
Line.
Specific operation process is that the bianry image is divided into three parts, respectively left part, middle part according to X direction
And right part；If the quantity for being distributed in the midpoint of left part is greater than first threshold, according to the coordinate value at the midpoint of all straight lines, sieve
The abscissa maximum value at midpoint is selected, and the midpoint that the corresponding midpoint coordinates of abscissa maximum value is determined as dividing straight line is sat
Mark；If the quantity for being distributed in the midpoint of right part is greater than the first threshold, according to the coordinate value at the midpoint of all straight lines, screening
The abscissa minimum value at midpoint out, and the corresponding midpoint coordinates of abscissa minimum value is determined as to divide the midpoint coordinates of straight line.
The bianry image is divided into three parts, respectively left part, middle part and right part according to horizontal axis method；If distribution
It is greater than second threshold in the quantity at middle part and the midpoint of lower part and filters out midpoint then according to the coordinate value at the midpoint of all straight lines
Ordinate minimum value, and by the corresponding midpoint coordinates of ordinate minimum value be determined as divide straight line midpoint coordinates.According to institute
The midpoint coordinates of segmentation straight line is stated, the equation of segmentation straight line is calculated, corresponding straight line in figure is then found according to linear equation.
The binary image is handled according to findcontours () function, obtains the edge of all enclosed regions
Curve；Straight line will be divided to connect with the boundary curve of the segmentation immediate enclosed region of straight line endpoint, determine segmentation contour.
Near point closed contour region is stuffed entirely with as white (255), remaining area filling becomes at black (0)
One using straight line as the black white image in watershed.The image is negated and carries out or operate to take out with high-altitude scene image
Close shot region, the high-altitude scene image by the image and after obscuring carries out or operation can take out the distant view region after obscuring.
Straight line is not present when carrying out straight-line detection to the edge in bianry image according to Hough straight line transformation algorithm, then first
To original image by OpenCV_BM algorithm obtain depth map (width judged by gray value distance grayscale image, gray scale
Worth range is 0-255), then some discontinuous regions are connected by corrosion and expansion algorithm.Because apart from close
Be that gray value is low, distance it is remote then to be that gray value is high be arranged a threshold value, the gray value for the pixel for being higher than threshold value
Become 0, the gray value lower than the pixel of threshold value becomes 1, thus obtains a width binary map.Then circle part is exactly gray value
For 0 part, close shot is exactly the region that gray value is 1.It distinguishes distant view and close shot is exactly to obscure to circle part later
With the reservation to close shot part.
Finally two figures are synthesized again, then available final image.
Compared with the prior art, the above method provided by the invention has very high real-time for the processing of video,
Simultaneously as the above method is based on computer vision, and run on the electronic equipment of the master controller with powerful performance
On, therefore personalized customization can be carried out to wearer, comparison optics subtract away from method more there is better picture to handle
The more high flexibility and height personalization suitability of effect and processing scheme, and algorithm improvement can be passed through in the later period, with
Extremely low cost increases substantially the treatment effect of this electronic glasses.
The present invention has higher because structurally having used modularization and readily removable design in hardware in maintenance
Maintainability, have lower maintenance cost.It is held high in addition, this electronic glasses for optical glasses, does not need replacement
Expensive optical mirror slip, equally also reduces manufacturing cost.
The present invention carries out image procossing for the scene that acrophobe visually sees, and is not treatment acrophobia
Shape, but alleviate acrophobia shape, to expand audient crowd.
Each embodiment in this specification is described in a progressive manner, the highlights of each of the examples are with other
The difference of embodiment, the same or similar parts in each embodiment may refer to each other.
Used herein a specific example illustrates the principle and implementation of the invention, and above embodiments are said
It is bright to be merely used to help understand method and its core concept of the invention；At the same time, for those skilled in the art, foundation
Thought of the invention, there will be changes in the specific implementation manner and application range.In conclusion the content of the present specification is not
It is interpreted as limitation of the present invention.
Claims (10)
1. a kind of electronic glasses, which is characterized in that including glasses box and binocular camera；The glasses box is embedded with successively
Liquid crystal display, master control borad and the mobile power source of setting；Master controller and six axis postures sensing are integrated on the master control borad
Device；The binocular camera is mounted on the lateral surface of the glasses box, the binocular camera, six axis posture sensing
Device, the mobile power source are electrically connected with the master controller.
2. electronic glasses according to claim 1, which is characterized in that be embedded at the wearing port of the glasses box
Sighting distance adjusts eyeglass, and the sighting distance adjusts the front that eyeglass is located at the liquid crystal display；The outside of the glasses box
Face is additionally provided with sighting distance and adjusts sliding shoe.
3. electronic glasses according to claim 2, which is characterized in that the wearing port of the glasses box is arc knot
Structure.
4. electronic glasses according to claim 1, which is characterized in that the glasses box assembled using clamping plate mode and
At the material of the glasses box is PLA material.
5. electronic glasses according to claim 1, which is characterized in that the master controller is reached using base frequency
Tetra- core processor of Intel Cherry Trail Z8350 of 1.44GHz；The master controller is integrated with HDMI video output end
Mouth and three USB ports；The master controller carries the Ubuntu system of open source and supports OpenCV_BM program.
6. electronic glasses according to claim 1, which is characterized in that the binocular camera is by two camera module groups
At the binocular camera is exported using single USB, and compatible Windows system and linux system；The binocular camera
Video output formats be yuv format.
7. electronic glasses according to claim 1, which is characterized in that the six axis attitude transducer is gyroscope acceleration
Meter.
8. electronic glasses according to claim 7, which is characterized in that the model MPU6050 of the gyroscope accelerometer.
9. a kind of high-altitude scene image method for reconstructing applied to electronic glasses described in claim 1-8 any one, special
Sign is that the high-altitude scene image method for reconstructing includes:
The high-altitude scene image of binocular camera acquisition and the head pose information of six axis attitude transducers acquisition are obtained, and really
Head rotation angle in the fixed head pose information；
Judge that the head rotation angle whether in set angle section, obtains the first judging result；
If first judging result indicates the head rotation angle not in set angle section, then by the high-altitude scape
As image is shown on liquid crystal display；
If first judging result indicates the head rotation angle in set angle section, then by the high-altitude scene
Image is converted into bianry image, and carries out edge detection to each region in the bianry image by Canny algorithm, obtains
The boundary curve of all areas；
All boundary curves are detected using Hough straight line transformation algorithm, straight line is judged whether there is, obtains second
Judging result；
If second judging result indicates that there are straight lines in the boundary curve, calculate the midpoint of all straight lines, then root
Determine segmentation contour according to the distributing position at the midpoint of all straight lines, and according to the segmentation contour to the high-altitude scene image into
Row segmentation, obtains close shot image and distant view image；
If second judging result indicates that straight line is not present in the boundary curve, using OpenCV_BM algorithm to described
High-altitude scene image is handled to obtain depth map, then according to erosion algorithm and expansion algorithm to the depth map at
Reason obtains the grayscale image of different depth, and according to the gray threshold of setting, carries out binaryzation division to the grayscale image, obtain
Close shot image and distant view image；The distant view image is gray value corresponding high-altitude scene image region when being 0, the close shot figure
Corresponding high-altitude scene image region when as be gray value being 1；Wherein, the gray value that will be above the pixel of the gray threshold becomes
At 0, the gray value that will be less than the pixel of the gray threshold becomes 1.
Fuzzy Processing is carried out to the distant view image；
By the distant view image synthesis after the close shot image and Fuzzy Processing, and composograph is shown on liquid crystal display.
10. scene image method for reconstructing in high-altitude according to claim 9, which is characterized in that described according to all straight lines
The distributing position at midpoint determines segmentation contour, specifically includes:
The bianry image is divided into three parts, respectively left part, middle part and right part according to X direction；If being distributed in a left side
The quantity at the midpoint in portion is greater than first threshold and filters out the abscissa at midpoint most then according to the coordinate value at the midpoint of all straight lines
Big value, and the corresponding midpoint coordinates of abscissa maximum value is determined as to divide the midpoint coordinates of straight line；If being distributed in right part
The quantity of point is greater than the first threshold, then according to the coordinate value at the midpoint of all straight lines, the abscissa for filtering out midpoint is minimum
Value, and the corresponding midpoint coordinates of abscissa minimum value is determined as to divide the midpoint coordinates of straight line；
The bianry image is divided into three parts, respectively left part, middle part and right part according to horizontal axis method；If in being distributed in
The quantity at the midpoint of portion and lower part is greater than second threshold, then according to the coordinate value at the midpoint of all straight lines, filters out the vertical of midpoint
Coordinate minimum value, and the corresponding midpoint coordinates of ordinate minimum value is determined as to divide the midpoint coordinates of straight line；
According to the midpoint coordinates of the segmentation straight line, segmentation straight line is calculated；
The binary image is handled according to findcontours () function, the edge for obtaining all enclosed regions is bent
Line；
Straight line will be divided to connect with the boundary curve of the segmentation immediate enclosed region of straight line endpoint, determine segmentation contour.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910095865.3A CN109829927B (en) | 2019-01-31 | 2019-01-31 | Electronic glasses and high-altitude scene image reconstruction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910095865.3A CN109829927B (en) | 2019-01-31 | 2019-01-31 | Electronic glasses and high-altitude scene image reconstruction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109829927A true CN109829927A (en) | 2019-05-31 |
| CN109829927B CN109829927B (en) | 2020-09-01 |
Family
ID=66863280
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910095865.3A Active CN109829927B (en) | 2019-01-31 | 2019-01-31 | Electronic glasses and high-altitude scene image reconstruction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109829927B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111552076A (en) * | 2020-05-13 | 2020-08-18 | 歌尔科技有限公司 | Image display method, AR glasses and storage medium |
| CN113113128A (en) * | 2021-04-15 | 2021-07-13 | 王小娟 | Medical operation auxiliary system and method based on VR, algorithm and 5G technology |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3492151B2 (en) * | 1996-09-10 | 2004-02-03 | 大日本スクリーン製造株式会社 | Image sharpness processing method and apparatus, and storage medium storing program |
| CN201662682U (en) * | 2010-04-27 | 2010-12-01 | 中兴通讯股份有限公司 | Stereo pick-up device |
| CN103297703A (en) * | 2013-05-30 | 2013-09-11 | 深圳职业技术学院 | Multispectral image acquisition method |
| CN105183170A (en) * | 2015-09-22 | 2015-12-23 | 京东方科技集团股份有限公司 | Head-wearing-type wearable equipment and information processing method and device thereof |
| CN105717644A (en) * | 2016-04-26 | 2016-06-29 | 上海与德通讯技术有限公司 | Image adjusting method and device and applicable intelligent glasses |
| CN106843491A (en) * | 2017-02-04 | 2017-06-13 | 上海肇观电子科技有限公司 | Smart machine and electronic equipment with augmented reality |
| CN107229340A (en) * | 2017-06-29 | 2017-10-03 | 联想(北京)有限公司 | A kind of information processing method and electronic equipment |
| CN107749053A (en) * | 2017-10-24 | 2018-03-02 | 郑州布恩科技有限公司 | A kind of binocular image collection and pretreatment unit and method for vision prosthesis |
-
2019
- 2019-01-31 CN CN201910095865.3A patent/CN109829927B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3492151B2 (en) * | 1996-09-10 | 2004-02-03 | 大日本スクリーン製造株式会社 | Image sharpness processing method and apparatus, and storage medium storing program |
| CN201662682U (en) * | 2010-04-27 | 2010-12-01 | 中兴通讯股份有限公司 | Stereo pick-up device |
| CN103297703A (en) * | 2013-05-30 | 2013-09-11 | 深圳职业技术学院 | Multispectral image acquisition method |
| CN105183170A (en) * | 2015-09-22 | 2015-12-23 | 京东方科技集团股份有限公司 | Head-wearing-type wearable equipment and information processing method and device thereof |
| CN105717644A (en) * | 2016-04-26 | 2016-06-29 | 上海与德通讯技术有限公司 | Image adjusting method and device and applicable intelligent glasses |
| CN106843491A (en) * | 2017-02-04 | 2017-06-13 | 上海肇观电子科技有限公司 | Smart machine and electronic equipment with augmented reality |
| CN107229340A (en) * | 2017-06-29 | 2017-10-03 | 联想(北京)有限公司 | A kind of information processing method and electronic equipment |
| CN107749053A (en) * | 2017-10-24 | 2018-03-02 | 郑州布恩科技有限公司 | A kind of binocular image collection and pretreatment unit and method for vision prosthesis |
Non-Patent Citations (2)
| Title |
|---|
| GABRIELE PETERS，JOCHEN KERDELS: ""Image Segmentation Based on Height Maps"", 《INTERNATIONAL CONFERENCE ON COMPUTER ANALYSIS OF IMAGES AND PATTERNS》 * |
| 茅天诒，王敬东，孙震，王崟: ""天空区域分割的暗原色先验去雾视觉优化技术"", 《光电技术应用》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111552076A (en) * | 2020-05-13 | 2020-08-18 | 歌尔科技有限公司 | Image display method, AR glasses and storage medium |
| CN113113128A (en) * | 2021-04-15 | 2021-07-13 | 王小娟 | Medical operation auxiliary system and method based on VR, algorithm and 5G technology |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109829927B (en) | 2020-09-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11281288B2 (en) | Eye and head tracking | |
| CN111415422B (en) | Virtual object adjustment method and device, storage medium and augmented reality equipment | |
| CN109993115B (en) | Image processing method and device and wearable device | |
| US20230360254A1 (en) | Pose estimation method and related apparatus | |
| CN108235774B (en) | Information processing method and device, cloud processing equipment and computer program product | |
| US20110148868A1 (en) | Apparatus and method for reconstructing three-dimensional face avatar through stereo vision and face detection | |
| CN109074681A (en) | Information processing unit, information processing method and program | |
| CN106598252A (en) | Image display adjustment method and apparatus, storage medium and electronic device | |
| US11843758B2 (en) | Creation and user interactions with three-dimensional wallpaper on computing devices | |
| JP7176520B2 (en) | Information processing device, information processing method and program | |
| US11589024B2 (en) | Multi-dimensional rendering | |
| US11699259B2 (en) | Stylized image painting | |
| CN105739707A (en) | Electronic equipment, face identifying and tracking method and three-dimensional display method | |
| CN109951698A (en) | For detecting the device and method of reflection | |
| CN111738914A (en) | Image processing method, image processing device, computer equipment and storage medium | |
| CN109829927A (en) | A kind of electronic glasses and high-altitude scene image method for reconstructing | |
| CN108064447A (en) | Method for displaying image, intelligent glasses and storage medium | |
| EP4164565A1 (en) | Blind assist eyewear with geometric hazard detection | |
| JP7400721B2 (en) | Information processing device, information processing method and program | |
| US20220030204A1 (en) | Artistic and other photo filter light field effects for images and videos utilizing image disparity | |
| CN110675413A (en) | Three-dimensional face model construction method and device, computer equipment and storage medium | |
| US11682368B1 (en) | Method of operating a mobile device | |
| Balakrishnan et al. | Stereopsis method for visually impaired to identify obstacles based on distance | |
| CN113327277A (en) | Three-dimensional reconstruction method and device for half-body image | |
| KR100651104B1 (en) | Gaze-based computer interface apparatus and method of using the same |
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 | ||
| GR01 | Patent grant |