CN110736965A - two-dimensional coding and decoding method for visible light positioning - Google Patents

Abstract

The invention discloses visible light positioning two-dimensional coding and decoding methods, which are OOK-based visible light coding and can be used for realizing indoor visible light positioning.A light source is coded and modulated by carrier frequency, so that the light source displays a bright-dark alternating polygonal pattern on a CMOS camera with fixed scanning frequency while illuminating.

Description

two-dimensional coding and decoding method for visible light positioning

Technical Field

The invention relates to two-dimensional coding and decoding methods for visible light positioning, belonging to the technical field of image processing and visible light positioning.

Background

With the rapid development of wireless sensor networks and internet of things technologies, is widely applied to the fields of intelligent robots, shopping guide in large shopping malls and the like, although indoor positioning technologies based on Bluetooth, radio frequency, ultrasonic waves, infrared rays, ultra-wide bandwidth, WiFi and the like are developed, the indoor positioning technologies need additional auxiliary positioning equipment, the indoor positioning cost is greatly increased, and the requirements on application environments are strict.

The principle of visible light positioning is that the high-speed flicker of the light emitting device is utilized to realize the positioning of the mobile equipment, the light emitting is represented as 1, the light is represented as 0, the light emitting device is modulated and encoded to emit a light signal with fixed information, the mobile equipment receives the light signal and obtains the position coordinate of the light source through decoding, and finally the coordinate of the mobile equipment is solved according to the geometric relationship between the light source and the mobile equipment, so that the positioning of the equipment is realized.

The existing visible light positioning system and method using an image sensor, such as ' indoor visible light self-positioning systems and methods based on binocular technology' (2018107961366) and 'positioning navigation system based on indoor lighting' (201210484538.5), 'binocular non-calibration space positioning method' (CN102622767B), ' indoor visible light positioning system' (CN107703485A) and ' indoor visible light asynchronous positioning method using a camera' (CN105306141A) all have the problems of single property of -dimensional coding, difficult identification and the like, are large in positioning error, low in positioning efficiency and easy to be influenced by indoor environmental factors, and have more or less defects in the aspects of accuracy, reliability, stability, universality and the like.

Therefore, a two-dimensional coding and decoding mode is needed to be adopted, so that the fault tolerance rate of visible light decoding is high, the effects of high decoding reliability and accurate result are achieved by obtaining multiple groups of data and eliminating abnormal numerical values in the data, meanwhile, the two-dimensional coding of the light source based on OOK is adopted, the realization of low cost is facilitated, and the method can be widely applied to indoor visible light positioning by pushing through .

Disclosure of Invention

The invention relates to a two-dimensional coding and decoding method for positioning visible light, which aims at the problems of the background technology, designs two-dimensional coding and decoding methods for positioning visible light, is used for realizing indoor visible light positioning, enables a light-emitting device to emit light at higher frequency by coding a light source and modulating the light source by determined carrier frequency, acquires images by a binocular camera fixed on equipment, processes and decodes the acquired images to obtain information carried by the images so as to distinguish the light source at different positions, and finally obtains the positioning of mobile equipment by calculating the binocular positioning technology.

two-dimensional coding and decoding method of visible light positioning, which is used to realize the coding of the light emitting device, is OOK-based visible light coding and modulates the visible light coding with the set carrier frequency, so that the light emitting device emits light with higher frequency, and the light source displays alternating bright and dark polygon patterns on the CMOS camera with fixed scanning frequency while illuminating, thereby the emitted light carries fixed information, and the bright and dark changes of the light source can not be perceived by human eyes in the working space of the light source;

the binocular camera is used for collecting images, and is two CMOS (Complementary Metal-Oxide-Semiconductor) cameras fixed on a certain plane of the mobile equipment, wherein the internal and external parameters of the cameras are the same, the imaging co-plane is formed, the optical axes are parallel to each other, and the scanning frequency is adjustable;

the mobile equipment is used for storing the mapping relation between the OOK codes of the light sources and the coordinates of the OOK codes, processing the images acquired by the binocular camera, distinguishing different light sources on the images according to a decoding method, obtaining the space coordinates of the light sources according to the mapping relation, and resolving the self-positioning of the mobile equipment by using a binocular positioning technology;

the decoding method is used for obtaining the coding information of the light source from the collected image so as to distinguish different light sources, the combined polygon is obtained after the collected image is processed by the mobile equipment, because the distance between the adjacent dark bands of each sub-area pattern is the same, the distance between the adjacent dark bands of different sub-area patterns is different, the ratio of the maximum polygon side length in the sub-area pattern to the average distance between the adjacent dark bands is calculated, and the light sources with different coordinates can be distinguished and the coordinates of the light sources can be obtained by sequencing the light sources according to the ratio.

, the specific implementation process of the encoding method includes the following steps:

step 1, OOK coding is carried out on a light source:

setting the scanning frequency of the binocular camera to be f_pResolution of the image being b₁×b₂OOK coding of each light source in lighting periods comprises hxb₁(h is a positive integer) 0-1 codes to ensure that there are at least complete patterns on the image for each light source, eachHxb formed by OOK coding₁All the 0 codes in the two-dimensional matrix at least comprise three same combined 6N polygons (N is a positive integer), all the centroids of the 6N polygons in each combined graph are the same, and the diagonals coincide, so that the code of the ith light source is as follows:wherein: b is₁、b₂Respectively the height and width of the image resolution, and the unit is pixel; h₁、H₂、H₃、M_iAre all composed of 0-1 code, wherein H₁、H₂、H₃A matrix consisting of all 1 codes is represented,a combination pattern representing the ith LED; symbol

Representing the kronecker product.

Step 2, calculating the maximum side length:

obtaining the side length c of the corresponding maximum polygon in the ith light source by combining multiple graphs in the step 1_iAnd a distance v_iRatio of (gamma)_i. According to gamma_iAnd numbering the light sources from 1 to 3 in sequence from small to large, recording the corresponding relation between the number of the light sources and the coordinates of the light sources, and storing the corresponding relation into the mobile equipment.

And 3, light source modulation:

with f_s＝f_p(hn) (n is a positive integer) such that the light source emits light at a higher frequency, wherein: f. of_sRepresenting the carrier frequency, f_pRepresenting the scan frequency, h represents the number of rows of the LED two-dimensional code in cycles.

Further , the specific implementation process of the decoding method includes the following steps:

step 1, image processing:

and simultaneously converting the two images into gray images, performing fuzzification processing on the gray images, and then enabling the gray images to pass through unilateral OTSU filters to find out the corresponding subarea of each light source on the images and the complete closed combined graph in the subarea.

Step 2, calculating the maximum side length:

and obtaining the side length of each side of a large polygon in the closed graph in the mth image according to the mth sub-area. Selecting any edge of the maximum polygon in the mth sub-region as a reference edge, and respectively recording 6N edges as the 1 st, 2 nd, N. The length of the N (N is more than or equal to 1 and less than or equal to 6N) th strip is marked as l_m,nRemoving abnormal values in the side length, namely the maximum value and the minimum value, obtaining the side length value with smaller error and solving the average value a of the residual side length_m。

And 3, calculating the side length distance:

and obtaining the adjacent side distance of two adjacent polygons in the mth image according to the mth sub-area. If the mth sub-region contains Y polygons, respectively recording the polygons as 1, 2, a_m,y,n. Averaging the distance D of the mth sub-region_m：

And 4, finding a corresponding light source code:

calculating the average value a of the side lengths of the maximum polygons in the mth sub-region_mAnd the distance mean value D_mRatio of (p)_m：

Pressing the light source at rho_mThe values are coded as 1, 2, 3 in descending order, respectively, by searching for codes previously stored in the deviceAnd mapping between the coordinate and the coordinate, and finding the coordinate value of the light source in the space in the image.

Compared with the existing indoor visible light coding technology, the OOK-based visible light coding and decoding technology has the following beneficial effects:

⑴ two-dimensional coding of the light source is realized based on OOK, the cost is low, the realization is easy, and the method can be widely used for indoor visible light positioning by ;

⑵, the two-dimensional coding is adopted, so that the visible light decoding fault-tolerant rate is high, errors are reduced by acquiring a plurality of groups of data and eliminating abnormal values therein, the decoding reliability is high, and the result is accurate;

⑶ the invention solves the problems of difficult recognition, large error and easy influence of indoor environment factors due to single of dimension code in visible light positioning using an image sensor, and has advantages in accuracy, reliability, stability and universality.

Drawings

FIG. 1 is a flow chart illustrating methods for implementing visible light positioning in a room according to the present invention;

FIG. 2 is a flow chart of OOK-based LED two-dimensional encoding methods according to the present invention;

FIG. 3 is a flow chart illustrating an OOK-based LED decoding method according to the present invention;

FIG. 4 shows an LED in the embodiment #1 of the present invention₁Partial coding;

FIG. 5 shows an LED in the embodiment #1 of the present invention₂Partial coding;

FIG. 6 shows an LED in the embodiment #1 of the present invention₃Partial coding;

FIG. 7 shows an LED in the embodiment #2 of the present invention₁Partial coding;

FIG. 8 shows an LED in the embodiment #2 of the present invention₂Partial coding;

FIG. 9 shows an LED in the embodiment #2 of the present invention₃And (4) partial coding.

Detailed Description

The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention.

FIG. 1 is a flow chart of a method for realizing indoor visible light positioning according to the present invention, which is functionally used for realizing encoding of a light emitting device, and is OOK-based visible light encoding, and the OOK-based visible light encoding is modulated by a set carrier frequency, so that the light emitting device emits light at a higher frequency, and a light source displays a light and dark alternating polygon pattern on a CMOS camera with a set scanning frequency of while illuminating, so that the emitted light carries set information, and human eyes cannot perceive the light and dark changes of the light source in a working space of the light source, and images are collected by a binocular camera fixed on a mobile device, and the obtained images are processed and decoded to obtain information carrying the light source to distinguish different positions, and finally the positioning of the mobile device is calculated by a binocular positioning technology.

The binocular camera is used for collecting images, and is two CMOS (Complementary Metal-Oxide-Semiconductor) cameras fixed on a certain plane of the mobile equipment, wherein the internal and external parameters of the cameras are the same, the imaging co-plane is formed, the optical axes are parallel to each other, and the scanning frequency is adjustable.

The mobile equipment is used for storing the mapping relation between the OOK codes of the light sources and the coordinates of the OOK codes, processing the images collected by the binocular camera, distinguishing different light sources on the images according to a decoding method, obtaining the space coordinates of the light sources according to the mapping relation, and resolving the self-positioning of the mobile equipment by using a binocular positioning technology.

The decoding method is used for obtaining the coding information of the light source from the collected image so as to distinguish different light sources, the combined polygon is obtained after the collected image is processed by the mobile equipment, because the distance between the adjacent dark bands of each sub-area pattern is the same, the distance between the adjacent dark bands of different sub-area patterns is different, the ratio of the maximum polygon side length in the sub-area pattern to the average distance between the adjacent dark bands is calculated, and the light sources with different coordinates can be distinguished and the coordinates of the light sources can be obtained by sequencing the light sources according to the ratio.

The core technical idea of the visible light positioning two-dimensional coding and decoding method is to realize new visible light coding and decoding, and in order to make the technical scheme of the invention clearer, the invention is described in detail in with reference to the attached drawings.

Fig. 2 is a flow chart of visible light positioning two-dimensional encoding and decoding methods in the present invention, implementing binary encoding function for light emitting devices as shown in fig. 1, which is OOK-based visible light encoding, so that a light source displays alternating bright and dark polygon patterns on a CMOS camera set with fixed scanning frequency while illuminating, so that the emitted light carries fixed information.

Step 1, OOK coding is carried out on a light source:

setting the scanning frequency of the binocular camera to be f_pResolution of the image being b₁×b₂OOK coding of each light source in lighting periods comprises hxb₁(h is a positive integer) 0-1 codes, each OOK code constitutes hxb to ensure that at least complete patterns exist on the image for each light source₁All the 0 codes in the two-dimensional matrix at least comprise three same combined 6N polygons (N is a positive integer), all the centroids of the 6N polygons in each combined graph are the same, and the diagonals coincide, so that the code of the ith light source is as follows:

wherein: b is₁、b₂Respectively the height and width of the image resolution, and the unit is pixel; h₁、H₂、H₃、M_iAre all composed of 0-1 code, wherein H₁、H₂、H₃A matrix representing the composition of all 1 codes,

a combination pattern representing the ith LED; symbol

Representing the kronecker product.

Step 2, calculating the maximum side length:

obtaining the side length c of the corresponding maximum polygon in the ith light source by combining multiple graphs in the step 1_iAnd a distance v_iRatio of (gamma)_i. According to gamma_iAnd numbering the light sources from 1 to 3 in sequence from small to large, recording the corresponding relation between the LED numbers and the coordinates of the LED numbers and storing the LED numbers and the coordinates into equipment.

And 3, light source modulation:

with f_s＝f_p(hn) (n is a positive integer) such that the light source emits light at a higher frequency, wherein: f. of_sRepresenting the carrier frequency, f_pRepresenting the scan frequency, h represents the number of rows of the LED two-dimensional code in cycles.

FIG. 3 is a flow chart of a two-dimensional encoding and decoding method for visible light positioning according to the present invention, which implements the binary decoding function of the light source shown in FIG. 1, for obtaining the encoded information of the light source from the collected image to distinguish different light sources.

Specifically, the specific implementation process of the decoding method comprises the following steps:

step 1, image processing:

and simultaneously converting the two images into gray images, performing fuzzification processing on the gray images, and then enabling the gray images to pass through unilateral OTSU filters to find out the corresponding subarea of each light source on the images and the complete closed combined graph in the subarea.

Step 2, calculating the maximum side length:

and obtaining the side length of each side of a large polygon in the closed graph in the mth image according to the mth sub-area. Selecting any edge of the maximum polygon in the mth sub-region as a reference edge, starting from the reference edge in the clockwise direction, respectively recording 6N edges as 1, 2,.. multidot.n,.. multidot.6N edges, and recording the N (N is more than or equal to 1 and less than or equal to 6N) edge as l_m,nRemoving abnormal values in the side length, namely the maximum value and the minimum value, obtaining the side length value with smaller error and solving the average value a of the residual side length_m。

And 3, calculating the side length distance:

and obtaining the adjacent side distance of two adjacent polygons in the mth image according to the mth sub-area. If the mth sub-region contains Y polygons, respectively recording the polygons as 1, 2, a_m,y,n. Averaging the distance D of the mth sub-region_m：

And 4, finding a corresponding light source code:

calculating the average value a of the side lengths of the maximum polygons in the mth sub-region_mAnd the distance mean value D_mRatio of (p)_m：

Pressing the light source at rho_mThe values are respectively coded as 1, 2 and 3 from small to large, and coordinate values of the light source in the image in the space are found by searching mapping between codes and coordinates which are stored in the device in advance.

The two-dimensional encoding and decoding methods for visible light positioning according to the present invention are further described in by two specific embodiments.

Example # 1:

setting scanning frequency f of binocular camera_p15.6kHz, lens focal length f_d4mm, the image resolution is 1280 × 960, and OOK encoding of lighting periods for each luminaire comprises 312 × 1280 0-1 codes, as shown in fig. 4, M is specified in the embodiment of the present invention₁、M₂、M₃Each of which is a matrix composed of a plurality of 0-1 codes, and in this embodiment #1, M is set₁The ith light source is denoted as L, where 39 × 40 matrices are composed of 0-1 codes_iThen is obtained by

The resulting matrix is L₁OOK encoding over cycles, where H₁156X 80 matrix, H, representing all elements 1₂156X 480 matrix, H, representing all elements 1₃4 x 4 matrix, symbol, representing all elements 1

Representing the kronecker product.

As shown in FIG. 5, setting M₂39 × 40, with M₁Same, then

The resulting matrix is L₂OOK encoding.

As shown in FIG. 6, setting M₃39 × 40, also with M₁Same, then

The resulting matrix is L₃OOK encoding. All the 0 codes in the two-dimensional matrix formed by each OOK code comprise three same combined positive 6-sided polygons, all the positive 6-sided polygons in each combined graph have the same centroids, and the diagonals are superposed.

By a matrixThe graph composed of the middle 0 code obtains L₁、L₂、L₃The length of the side of the maximum hexagon in the two-dimensional code is as follows: c. C₁＝c₂＝c₃Distance between adjacent sides, 68: v. of₁＝8，v₂＝16，v₃24, side length to distance ratio: gamma ray₁＝17/2，γ₂＝17/4，γ₃17/6. According to gamma_iAnd respectively numbering the light sources from 1 to 3 again from small to large, recording the corresponding relation between the numbers and the coordinates of the light sources and storing the numbers and the coordinates into the mobile equipment.

The carrier frequency f is obtained by h 312 and n1₁50 Hz. By OOK-encoding the light source and using the carrier frequency f_sThe method comprises the steps of modulating a light source at 110Hz to enable the light source to emit light at a high frequency, collecting images of the light source through a binocular camera at a moving equipment end, processing the captured images, increasing the contrast of the captured images to enhance the edge characteristic, simultaneously converting the two images into gray-scale images, performing fuzzification processing on the gray-scale images, enabling the gray-scale images to pass through single-side OTSU filters, and finding corresponding sub-areas of each light source on the images and closed graphs in the sub-areas.

Selecting any side of the maximum polygon in the mth sub-region as a reference side, starting from the reference side in the clockwise direction, respectively marking 6 sides as 1, 2, 3, 4, 5 and 6 sides, and marking the side length of the nth (n is not less than 1 and not more than 6) as l_m,n，

l_1,1＝68，l_1,2＝67，l_1,3＝69，l_1,4＝68，l_1,5＝67，l_1,6＝68，

l_2,1＝68，l_2,2＝66，l_2,3＝70，l_2,4＝68，l_2,5＝67，l_2,6＝67，

l_3,1＝67，l_3,2＝68，l_3,3＝70，l_3,4＝68，l_3,5＝68，l_3,6＝68。

Eliminating abnormal value l in side length_1,2，l_1,3，l_2,2，l_2,3，l_3,1，l_3,3I.e. the maximum value and the minimum value, obtaining the edge length value with smaller error and solvingMean value of remaining edge length a_m,

a₁＝68，a₂＝68，a₃＝68

And obtaining the adjacent side distance of two adjacent polygons in the mth image according to the mth sub-area. The 1 st sub-area contains 3 polygons, the polygons are respectively marked as 1, 2 and 3 according to the sequence from large to small, the 2 nd sub-area contains 5 polygons, the polygons are respectively marked as 1, 2, 5 according to the sequence from large to small, the 3 rd sub-area contains 2 polygons, and the polygons are respectively marked as 1 and 2 according to the sequence from large to small. The numbers of adjacent sides of the (y + 1) th polygon and the (y + 1) th polygon are the same, and the distance d between the nth side of the (y) th polygon and the nth side of the (y + 1) th polygon in the (m) th sub-region is obtained_m,y,n。

th subregion:

d_1,1,1＝15，d_1,1,2＝15，d_1,1,3＝17，d_1,1,4＝16，d_1,1,5＝16，d_1,1,6＝17，

d_1,2,1＝17，d_1,2,2＝16，d_1,2,3＝17，d_1,2,4＝16，d_1,2,5＝18，d_1,2,6＝17，

in the second subregion:

d_2,1,1＝8，d_2,1,2＝9，d_2,1,3＝8，d_2,1,4＝9，d_2,1,5＝9，d_2,1,6＝8，

d_2,2,1＝9，d_2,2,2＝9，d_2,2,3＝7，d_2,2,4＝7，d_2,2,5＝9，d_2,2,6＝7，

d_2,3,1＝7，d_2,3,2＝8，d_2,3,3＝9，d_2,3,4＝9，d_2,3,5＝7，d_2,3,6＝8，

d_2,4,1＝8，d_2,4,2＝9，d_2,4,3＝7，d_2,4,4＝9，d_2,4,5＝9，d_2,4,6＝8，

in the third subregion:

d_3,1,1＝25，d_3,1,2＝23，d_3,1,3＝24，d_3,1,4＝24，d_3,1,5＝24，d_3,1,6＝25，

averaging the distance D of the mth sub-region_m：

D₁＝17，D₂＝8，D₃＝24

Calculating the average value a of the side lengths of the maximum polygons in the mth sub-region_mAnd the distance mean value D_mThe ratio of (A) to (B):

ρ₁＝4，ρ₂＝17/2，ρ₃＝17/6

mixing L with_iAccording to rho_mSequence pairs L with values from small to large₃、L₁、L₂The codes are 1, 2 and 3 respectively, and coordinate values of the light source in the image in the space are found by searching mapping between the codes and the coordinates which are stored in the mobile device in advance.

Example # 2:

setting scanning frequency f of binocular camera_p15.6kHz, lens focal length f_dThe image resolution is 1280 × 960 at 4mm, and OOK encoding of each light source over lighting periods comprises 608 × 1280 0-1 codes.

As shown in FIG. 7, T is specified in the embodiment of the present invention₁、T₂、T₃Each of which is a matrix composed of a plurality of 0-1 codes, and in this embodiment #2, T is set₁The ith light source is denoted as L, where 76 × 76 matrices of 0-1 codes are formed_iThen is obtained byThe resulting matrix is L₁OOK encoding over cycles, where W₁304 x 152 matrix representing all elements as 1，W₂304 x 336 matrix, W, representing all elements 1₃4 x 4 matrix, symbol, representing all elements 1

Representing the kronecker product.

As shown in FIG. 8, T is set₂76 × 76, and T₁Same, then

The resulting matrix is L₂OOK encoding.

As shown in FIG. 9, T is set₃76 × 76, also with T₁Same, then

The resulting matrix is L₃OOK encoding. All the 0 codes in the two-dimensional matrix formed by each OOK code comprise 3 same combined positive 12-sided polygons, all the positive 12-sided polygons in each combined graph have the same centroids and the diagonals are superposed.

By a matrix

The graph composed of the middle 0 code obtains L₁、L₂、L₃The two-dimensional codes of (2) correspond to the maximum dodecagon side length: c. C₁＝c₂＝c₃80, distance between adjacent sides: v. of₁＝16，v₂＝32，v₃48, side length to distance ratio: gamma ray₁＝5，γ₂＝5/2，γ₃5/3. According to gamma_iAnd respectively numbering the light sources from 1 to 3 again from small to large, recording the corresponding relation between the numbers and the coordinates of the light sources and storing the numbers and the coordinates into the mobile equipment.

By OOK-encoding the light source and using the carrier frequency f₁The light source is modulated at 50Hz so that the light source emits light at a higher frequency. The moving equipment end collects light source images through a binocular camera, processes the captured images, and enhances the edge characteristics by increasing the contrast of the captured images. Simultaneously combine twoAnd converting the image into a gray-scale image, performing fuzzification processing on the gray-scale image, and then enabling the gray-scale image to pass through unilateral OTSU filters to find out a corresponding sub-area of each light source on the image and a closed graph in the sub-area.

Selecting any side of the maximum polygon in the mth sub-region as a reference side, starting from the reference side in the clockwise direction, respectively marking 12 sides as 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 and 12 sides, and marking the side length of the nth (n is more than or equal to 1 and less than or equal to 12) as l_m,n：

l_1,1＝83，l_1,2＝77，l_1,3＝82，l_1,4＝82，l_1,5＝81，l_1,6＝79，

l_1,7＝80，l_1,8＝78，l_1,9＝80，l_1,10＝79，l_1,11＝79，l_1,12＝79，

l_2,1＝81，l_2,2＝79，l_2,3＝80，l_2,4＝81，l_2,5＝83，l_2,6＝82，

l_2,7＝79，l_2,8＝79，l_2,9＝77，l_2,10＝79，l_2,11＝81，l_2,12＝79，

l_3,1＝80，l_3,2＝76，l_3,3＝79，l_3,4＝79，l_3,5＝79，l_3,6＝79，

l_3,7＝80，l_3,8＝78，l_3,9＝82，l_3,10＝83，l_3,11＝80，l_3,12＝80，

Eliminating abnormal value l in side length_1,1，l_1,2，l_2,5，l_2,9，l_3,2，l_3,10I.e. the maximum value and the minimum value, obtaining the edge length value with smaller error and solving the average value a of the residual edge length_m：

a₁＝80，a₂＝80，a₃＝80

And obtaining the adjacent side distance of two adjacent polygons in the mth image according to the mth sub-area. The 1 st sub-region contains 3 polygons, i.e., a polygonThe shapes are respectively marked as 1, 2 and 3 from large to small, the 2 nd sub-area comprises 5 polygons, the polygons are respectively marked as 1, 2, 5 from large to small, the 3 rd sub-area comprises 2 polygons, and the polygons are respectively marked as 1 and 2 from large to small. The numbers of adjacent sides of the (y + 1) th polygon and the (y + 1) th polygon are the same, and the distance d between the nth side of the (y) th polygon and the nth side of the (y + 1) th polygon in the (m) th sub-region is obtained_m,y,n。

th subregion:

d_1,1,1＝48，d_1,1,2＝46，d_1,1,3＝47，d_1,1,4＝47，d_1,1,5＝49，d_1,1,6＝46，

d_1,1,7＝46，d_1,1,8＝47，d_1,1,9＝49，d_1,1,10＝49，d_1,1,11＝49，d_1,1,12＝48，

in the second subregion:

d_2,1,1＝32，d_2,1,2＝31，d_2,1,3＝32，d_2,1,4＝32，d_2,1,5＝33，d_2,1,6＝32，

d_2,1,7＝31，d_2,1,8＝34，d_2,1,9＝30，d_2,1,10＝34，d_2,1,11＝34，d_2,1,12＝33，

d_2,2,1＝30，d_2,2,2＝31，d_2,2,3＝31，d_2,2,4＝33，d_2,2,5＝31，d_2,2,6＝33，

d_2,2,7＝30，d_2,2,8＝33，d_2,2,9＝32，d_2,2,10＝33，d_2,2,11＝33，d_2,2,12＝34。

in the third subregion:

d_3,1,1＝17，d_3,1,2＝16，d_3,1,3＝15，d_3,1,4＝15，d_3,1,5＝18，d_3,1,6＝14，

d_3,1,7＝16，d_3,1,8＝15，d_3,1,9＝18，d_3,1,10＝17，d_3,1,11＝16，d_3,1,12＝16，

d_3,2,1＝14，d_3,2,2＝17，d_3,2,3＝14，d_3,2,4＝14，d_3,2,5＝16，d_3,2,6＝14，

d_3,2,7＝17，d_3,2,8＝17，d_3,2,9＝17，d_3,2,10＝15，d_3,2,11＝17，d_3,2,12＝16，

d_3,3,1＝18，d_3,3,2＝17，d_3,3,3＝17，d_3,3,4＝16，d_3,3,5＝16，d_3,3,6＝17，

d_3,3,7＝14，d_3,3,8＝15，d_3,3,9＝15，d_3,3,10＝16，d_3,3,11＝17，d_3,3,12＝17，

d_3,4,1＝14，d_3,4,2＝16，d_3,4,3＝16，d_3,4,4＝16，d_3,4,5＝17，d_3,4,6＝17，

d_3,4,7＝15，d_3,4,8＝16，d_3,4,9＝14，d_3,4,10＝18，d_3,4,11＝15，d_3,4,12＝14。

averaging the distance D of the mth sub-region_m：

D₁＝47，D₂＝32，D₃＝16

Calculating the average value a of the side lengths of the maximum polygons in the mth sub-region_mAnd the distance mean value D_mRatio of (p)_m：

ρ₁＝4，ρ₂＝17/2，ρ₃＝17/6

Mixing L with_iAccording to rho_mSequence pairs L with values from small to large₃、L₂、L₁The codes are 1, 2 and 3 respectively, and coordinate values of the light source in the image in the space are found by searching mapping between the codes and the coordinates which are stored in the mobile device in advance.

From the above results, it can be seen that the visible light positioning two-dimensional encoding and decoding methods in embodiments #1 and #2 can achieve encoding and decoding of the light source, and the obtained results have high efficiency, reliability and accuracy.

In summary, the two-dimensional coding and decoding methods for visible light positioning of the present invention utilize the fast switching characteristic of the light emitting device to realize communication by the high-speed flashing light thereof, thereby effectively solving the problem of unreliable communication in indoor visible light positioning.

It will be appreciated by those skilled in the art that modifications and variations can be made to the disclosed embodiments without departing from the spirit and scope of the invention, and therefore, is equivalent to modifications and variations that would be apparent to those skilled in the art without departing from the spirit and scope of the invention as disclosed in the appended claims.

Claims (3)

1. The method is characterized in that OOK-based visible light coding is performed, the OOK-based visible light coding is modulated by set carrier frequency, so that the light emitting device emits light at higher frequency, a light source displays a polygonal pattern with alternating light and shade on a CMOS camera with fixed scanning frequency while illuminating, the emitted light carries fixed information, human eyes cannot perceive the light and shade change of the light source in a working space of the light source, images are collected through a binocular camera fixed on mobile equipment, the obtained images are processed and decoded, information carried by the images is obtained to distinguish the light sources at different positions, and finally the positioning of the mobile equipment is obtained through calculation of a binocular positioning technology;

   the binocular camera is used for collecting images, the images are two CMOS cameras fixed on a certain plane of the mobile equipment, the internal and external parameters of the cameras are the same, the images are coplanar, the optical axes are parallel to each other, and the scanning frequency is adjustable;

   the mobile equipment is used for storing the mapping relation between the OOK codes of the light sources and the coordinates of the OOK codes, processing the images acquired by the binocular camera, distinguishing different light sources on the images according to a decoding method, obtaining the space coordinates of the light sources according to the mapping relation, and resolving the self-positioning of the mobile equipment by using a binocular positioning technology;

   the decoding method is used for obtaining the coding information of the light source from the collected image so as to distinguish different light sources, the combined polygon is obtained after the collected image is processed by the mobile equipment, because the distance between the adjacent dark bands of each sub-area pattern is the same, the distance between the adjacent dark bands of different sub-area patterns is different, the ratio of the maximum polygon side length in the sub-area pattern to the average distance between the adjacent dark bands is calculated, and the light sources with different coordinates can be distinguished and the coordinates of the light sources can be obtained by sequencing the light sources according to the ratio.
2. 2. The two-dimensional coding and decoding method for visible light positioning according to claim 1, wherein the two-dimensional coding method includes the following steps:

   step 1, OOK coding is carried out on a light source:

   setting the scanning frequency of the binocular camera to be f_pResolution of the image being b₁×b₂OOK coding of each light source in lighting periods comprises hxb₁0-1 code (h is positive integer), and h × b coded by OOK to ensure that each light source has at least complete patterns on the image₁All the 0 codes in the two-dimensional matrix at least comprise three same combined 6N polygons (N is a positive integer), all the centroids of the 6N polygons in each combined graph are the same, and the diagonals coincide, so that the code of the ith light source is as follows:

   b is₁、b₂Respectively the height and width of the image resolution, and the unit is pixel; h₁、H₂、H₃、M_iAre all composed of 0-1 code, wherein H₁、H₂、H₃A matrix consisting of all 1 codes is represented,

   

   a combination pattern representing the ith LED; symbol

   

   Represents the kronecker product;

   step 2, calculating the maximum side length:

   obtaining the side length c of the corresponding maximum polygon in the ith light source by combining multiple graphs in the step 1_iAnd a distance v_iRatio of (gamma)_i(ii) a According to gamma_iNumbering the light sources from 1 to 3 in sequence from small to large, recording the corresponding relation between the numbers of the light sources and the coordinates of the light sources, and storing the numbers into the mobile equipment;

   and 3, light source modulation:

   with f_s＝f_p(hn) (where n is a positive integer) such that the light source emits light at a higher frequency, wherein: f. of_sRepresenting the carrier frequency, f_pRepresenting the scan frequency, h represents the number of rows of the LED two-dimensional code in cycles.
3. 3. The two-dimensional coding and decoding method for visible light positioning as claimed in claim 1, wherein the decoding method comprises the following steps:

   step 1, image processing:

   simultaneously converting the two images into gray images, performing fuzzification processing on the gray images, and then enabling the gray images to pass through unilateral OTSU filters to find out corresponding subregions of each light source on the images and complete closed combined graphs in the subregions;

   step 2, calculating the maximum side length:

   obtaining the side length of each side of a large polygon in a closed graph in the mth image according to the mth sub-area; selecting any edge of the maximum polygon in the mth sub-region as a reference edge, and respectively recording 6N edges as the 1 st, 2 nd, N. The nth side (N is more than or equal to 1 and less than or equal to 6N) is recorded as l_m,nRemoving abnormal values in the side length, namely the maximum value and the minimum value, obtaining the side length value with smaller error and solving the average value a of the residual side length_m；

   And 3, calculating the side length distance:

   obtaining the adjacent side distance of two adjacent polygons in the mth image according to the mth sub-area; if the mth sub-region contains Y polygons, respectively recording the polygons as 1, 2, a_m,y,n(ii) a Averaging the distance D of the mth sub-region_m：

   

   And 4, finding a corresponding light source code:

   calculating the average value a of the side lengths of the maximum polygons in the mth sub-region_mAnd the distance mean value D_mRatio of (p)_m：

   

   Pressing the light source at rho_mThe values are coded as 1, 2 and 3 in descending order, and the mapping between the codes and the coordinates stored in the equipment in advance is searched to find outTo the coordinate values of the light source in space in the image.

Images