CN108801257A - A kind of localization method for indoor automatic parking - Google Patents

Abstract

The invention discloses a positioning method for indoor automatic parking. Enter the GIS database; (3) The camera takes pictures of the grid on the ground and uploads them to the server; (4) Identify the grid information and read out the binary number sequence; (5) Query the vehicle in the GIS database according to the output binary number sequence actual location. The invention combines technologies such as computer vision, location coding, and location services. After storing the location data in the database, the binary ID information is obtained by identifying the drawn grid. Vehicle positioning has a wide range of applications.

Description

A positioning method for indoor automatic parking

technical field

The invention relates to a vehicle parking method, in particular to a positioning method for indoor automatic parking. .

Background technique

With the continuous advancement of science and technology, indoor positioning technology continues to develop, including Bluetooth indoor positioning technology, infrared indoor positioning technology, ultra-wideband indoor positioning technology, Zigbee indoor positioning technology, WiFi indoor positioning technology, etc. However, these technologies are inconvenient to use and cannot meet the positioning requirements of automatic parking. Therefore, there is an urgent need for a stable, high-precision and easy-to-use positioning method for indoor automatic parking.

Contents of the invention

Purpose of the invention: In view of the problems existing in the prior art, the purpose of the present invention is to provide a positioning method for indoor automatic parking that is convenient to use and has high positioning accuracy.

Technical solution: a positioning method for indoor automatic parking, comprising the following steps:

(1) Spray the grid on the road of the indoor parking lot, and the grid adopts binary code;

(2) Enter the binary number sequence corresponding to the actual position into the GIS database;

(3) The camera takes pictures of the grid on the ground and uploads them to the server;

(4) Identify the grid information and read out the binary number sequence;

(5) Query the actual position of the vehicle in the GIS database according to the output binary number sequence.

Described step (1) comprises following content:

(1.1) Adopt n-bit binary code, the main part has n grids, and the size of a grid is a meter*a meter, then the road width that can be represented is n*a=na meter, and the road length that can be represented is 2 ⁿ * a=2 ⁿ a meters, the representable area is na*2 ⁿ a=n2 ⁿ a ² square meters;

(1.2) The grid is sprayed in black and white, black represents the binary number 1, and white represents the binary number 0;

(1.3) Set the start bit, stop bit and parity bit; the start bit is black, set to "1", the stop bit is white, set to "0"; use odd or even parity bit, so that the start bit The number of "1" between the stop bit and the stop bit is always odd or even;

(1.4) Sequential coding, the ID information of each road section is continuous from the beginning to the end.

Described step (2) comprises following content:

In the GIS database, the straight-line road section is represented by its type, the binary ID information at both ends and their actual coordinates: {S; ID1, (X1, Y1); ID2, (X2, Y2)}; the circular arc road section is represented by its Type, angle, center coordinates and binary ID information at both ends and their actual coordinate representation: {C; θ; (X0, Y0); ID1, (X1, Y1); ID2, (X2, Y2)}.

Described step (4) comprises following content:

(4.1) Grayscale the color grid image captured by the camera to reduce the amount of data, storage space and image processing time;

(4.2) Use the median filter to remove the noise points introduced when taking the grid image, and reduce the interference of noise on the subsequent image processing;

(4.3) Binarize the image, and use the algorithm combining adaptive illumination equalization and Otsu global threshold to binarize the grid image to remove the influence of uneven illumination of the grid image;

(4.4) Extract edge features and calculate the total length L, then the coordinates of the i-th sampling point are (Li/(n+3), a/2), i<=n+3;

(4.5) Sampling the sampling point according to the obtained coordinates, the dark color is recorded as "0", and the light color is recorded as "1";

(4.6) Output binary codewords from the start bit to the stop bit; regardless of whether the vehicle is driving forward or backward, the first sampling point and the last sampling point must be "0" and the other is "1"; if "1" is in At the left end, the binary code word is output from left to right, and after verification, the start bit, stop bit and check bit are discarded to obtain an ID information; if "1" is at the right end, the binary code word is output from right to left, and the calibration After verification, the start bit, stop bit and check bit are discarded to obtain an ID information.

Described step (5) comprises following content:

(5.1) According to the detected binary ID information, search for the ID between the two end IDs of which road section, then the actual position is on the road section;

(5.2) According to the data of the actual road section searched, if it is {S; ID1, (X1, Y1); ID2, (X2, Y2)}, that is, the road section is linear, convert (ID-ID1) into a decimal number A, convert (ID2-ID1) into decimal number B, then the current actual coordinates are (A(X2-X1)/B,A(Y2-Y1)/B); if the road section data is {C; θ; (X0 ,Y0); ID1, (X1,Y1); ID2,(X2,Y2)}, then the arc radius R＝√(X0-X1) ² +(Y0-Y1) ² , then the current actual coordinate is

Beneficial effect

Compared with the prior art, the present invention has the following remarkable progress: the present invention combines technologies such as computer vision, location coding and location services, and after storing the location data in the database, obtains the binary ID information by identifying the drawn grid, and uses Convenient, high positioning accuracy, accurate and reliable vehicle positioning indoors, wide application range.

Description of drawings

Fig. 1 is a schematic flow chart of the construction method of the present invention;

Fig. 2 is a schematic diagram of a grid of a straight road segment;

Fig. 3 is the schematic diagram of circular arc section;

Fig. 4 is a schematic diagram of calculating sampling points.

Detailed ways

The technical scheme of the present invention will be described in further detail below in conjunction with the implementation examples and the accompanying drawings.

As shown in Figure 1, a positioning method for indoor automatic parking includes the following steps:

Step 1: Spray the grid on the road of the indoor parking lot, and the grid adopts binary code. Using 32-bit binary code, there are 32 grids in the main part, and the size of a grid is 25cm*25cm, then the road width that can be represented is 32*0.25=8 meters, and the road length that can be represented is 2 ³² *0.25=2 ³⁰ meters , the representable area is 8*2 ³⁰ =2 ³³ square meters, which is about 8590 square kilometers. The grid is painted in black and white, with black representing the binary number 1 and white representing the binary number 0. Set start bit, stop bit and parity bit. The start bit is black and set to "1" and the stop bit is white and set to "0". The even parity bit is used so that the number of "1"s between the start bit and the stop bit is always an even number. Sequential coding, the ID information of each road segment is continuous from the beginning to the end.

Step 2, enter the binary number sequence corresponding to the actual location into the GIS database. In the GIS database, a straight road segment is represented by its type, the binary ID information of its two ends and their actual coordinates. In this example, the straight-line road section is expressed as {S; ID1, (X1, Y1); ID2, (X2, Y2)}; the arc road section is represented by its type, center coordinates, angle and binary ID information at both ends and their Actual coordinate representation. In this example, the arc segment is expressed as {C; θ; (X0, Y0); ID1, (X1, Y1); ID2, (X2, Y2)}.

Step 3, when the vehicle enters the indoor parking lot, the on-board camera takes pictures of the grid on the ground and uploads them to the server;

Step 4, identify the grid information and read out the sequence of binary numbers. The color of each pixel in a color image is determined by three components of R, G, and B, and each component has 255 values, so that a pixel can have more than 16 million (255*255*255) color changes scope. The variation range of one pixel of grayscale image is only 255 kinds, but the description of grayscale image still reflects the distribution and characteristics of the overall and local chromaticity and brightness level of the whole image just like color image. Grayscale the color grid image captured by the camera to reduce the amount of data, storage space and image processing time. Take an odd number from a sampling window in the image and arrange them in ascending or descending order, calculate the arithmetic average of the values of each point after sorting, and replace the value to be processed with the calculated median value, so that the surrounding pixel values are close to the real value, Realize the median filtering of the image, remove the noise points introduced when taking the grid image, and reduce the interference of noise on the subsequent image processing. The image is binarized, and the grid image is binarized by an algorithm combining adaptive illumination equalization and Otsu global threshold to remove the influence of uneven illumination of the grid image. The flow of the Otsu algorithm is as follows. Suppose the image contains L gray levels (0, 1..., L-1), the number of pixels with gray value i is Ni, and the total number of pixels in the image is N=N0+N1+...+ N(L-1). The generalization of the point with the gray value i is: P(i)=N(i)/N. The threshold t divides the whole image into dark area c1 and bright area c2, and the variance σ between classes is t Function: σ=a1*a2(u1-u2)^2(2) In the formula, aj is the ratio of the area of class cj and the total area of the image, a1=sum(P(i))i->t, a2= 1-a1; uj is the mean value of class cj, u1=sum(i*P(i))/a1 0->t, u2=sum(i*P(i))/a2,t+1->L- 1 This method selects the optimal threshold t^ to maximize the variance between classes, that is: let Δu=u1-u2, σb=max{a1(t)*a2(t)Δu^2}. Extract the edge features of the image, extract the grid from the entire image, and then calculate the total length L of the grid, then the coordinates of the i-th sampling point are (0.25/2, Li/34). The binarized grid image is sampled according to the obtained sampling point coordinates, and the dark color is recorded as "0", and the light color is recorded as "1". Output a binary codeword from the start bit to the stop bit. Regardless of whether the vehicle is driving forward or backward, the first sampling point and the last sampling point must be "0" and "1". If "1" is at the left end, the binary code word is output from left to right, and after verification, the start bit, stop bit and check bit are discarded to obtain an ID information; if "1" is at the right end, it is output from right to left Binary code word, after verification, the start bit, stop bit and check bit are discarded to obtain an ID information.

Step five, query the actual position of the vehicle in the GIS database according to the output binary number sequence. According to the detected binary ID information, search for the ID between the two end IDs of the road section, and the actual location is on the road section.在本实例中，假设检测到的ID为(01000000000000000000000000000000)，若搜索出的实际路段的数据为{S；ID1，(X1,Y1)；ID2,(X2,Y2)}，即路段为直线型， Convert (ID-ID1) to decimal number 2, and convert (ID2-ID1) to decimal number 11, then the current actual coordinates are (2(X2-X1)/11,2(Y2-Y1)/11); if The road section data is {C; θ; (X0, Y0); ID1, (X1, Y1); ID2, (X2, Y2)}, that is, the road section is arc-shaped, and the radius of the arc is R=√(X0-X1) ² +(Y0-Y1) ² , the current actual coordinate is

Claims (5)

1. A positioning method for indoor automatic parking, comprising the steps of: (1) Spray the grid on the road of the indoor parking lot, and the grid adopts binary code; (2) Enter the binary number sequence corresponding to the actual position into the GIS database; (3) The camera takes pictures of the grid on the ground and uploads them to the server; (4) Identify the grid information and read out the binary number sequence; (5) Query the actual position of the vehicle in the GIS database according to the output binary number sequence. 2. The positioning method for indoor automatic parking according to claim 1, characterized in that: said step (1) includes the following content: (1.1) Adopt n-bit binary code, the main part has n grids, and the size of a grid is a meter*a meter, then the road width that can be represented is n*a=na meter, and the road length that can be represented is 2 ⁿ * a=2 ⁿ a meters, the representable area is na*2 ⁿ a=n2 ⁿ a ² square meters; (1.2) The grid is sprayed in black and white, black represents the binary number 1, and white represents the binary number 0; (1.3) Set the start bit, stop bit and parity bit; the start bit is black, set to "1", the stop bit is white, set to "0"; use odd or even parity bit, so that the start bit The number of "1" between the stop bit and the stop bit is always odd or even; (1.4) Sequential coding, the ID information of each road section is continuous from the beginning to the end. 3. The positioning method for indoor automatic parking according to claim 2, characterized in that: said step (2) includes the following content: In the GIS database, the straight-line road section is represented by its type, the binary ID information at both ends and their actual coordinates: {S; ID1, (X1, Y1); ID2, (X2, Y2)}; the circular arc road section is represented by its Type, angle, center coordinates and binary ID information at both ends and their actual coordinate representation: {C; θ; (X0, Y0); ID1, (X1, Y1); ID2, (X2, Y2)}. 4. The positioning method for indoor automatic parking according to claim 3, characterized in that: said step (4) includes the following content: (4.1) Grayscale the color grid image captured by the camera to reduce the amount of data, storage space and image processing time; (4.2) Use the median filter to remove the noise points introduced when taking the grid image, and reduce the interference of noise on the subsequent image processing; (4.3) Binarize the image, and use the algorithm combining adaptive illumination equalization and Otsu global threshold to binarize the grid image to remove the influence of uneven illumination of the grid image; (4.4) Extract edge features and calculate the total length L, then the coordinates of the i-th sampling point are (Li/(n+3), a/2), i<=n+3; (4.5) Sampling the sampling point according to the obtained coordinates, the dark color is recorded as "0", and the light color is recorded as "1"; (4.6) Output binary codewords from the start bit to the stop bit; regardless of whether the vehicle is driving forward or backward, the first sampling point and the last sampling point must be "0" and the other is "1"; if "1" is in At the left end, the binary code word is output from left to right, and after verification, the start bit, stop bit and check bit are discarded to obtain an ID information; if "1" is at the right end, the binary code word is output from right to left, and the calibration After verification, the start bit, stop bit and check bit are discarded to obtain an ID information. 5. A positioning method for indoor automatic parking according to claim 4, characterized in that: said step (5) includes the following content: (5.1) According to the detected binary ID information, search for the ID between the two end IDs of which road section, then the actual position is on the road section; (5.2) According to the data of the actual road section searched, if it is {S; ID1, (X1, Y1); ID2, (X2, Y2)}, that is, the road section is linear, convert (ID-ID1) into a decimal number A, convert (ID2-ID1) into decimal number B, then the current actual coordinates are (A(X2-X1)/B,A(Y2-Y1)/B); if the road section data is {C; θ; (X0 ,Y0); ID1, (X1,Y1); ID2,(X2,Y2)}, then the arc radius R＝√(X0-X1) ² +(Y0-Y1) ² , then the current actual coordinate is