CN110471072B

CN110471072B - Method and system for identifying position of reflecting column

Info

Publication number: CN110471072B
Application number: CN201910762994.3A
Authority: CN
Inventors: 詹鹏飞; 王俊石; 娄兵兵
Original assignee: Qingdao Huashine Intelligent Technology Co ltd
Current assignee: Huasheng Qingdao Intelligent Equipment Technology Co ltd; Qingdao Huasheng Intelligent Equipment Co ltd
Priority date: 2019-08-19
Filing date: 2019-08-19
Publication date: 2021-04-02
Anticipated expiration: 2039-08-19
Also published as: CN110471072A

Abstract

The invention discloses a method and a system for identifying the position of a reflecting column, which relate to the field of intelligent warehouse logistics and mainly comprise the following steps: acquiring sampling point data of the reflective columns; the reflective column sampling point data comprises a plurality of groups of data points; each group of data points comprises sampling point light intensity data, sampling point distance data and sampling point angle data; sorting data points in the sampling point data of the reflective column according to the size of the sampling point angle data; selecting data points of which the distance data of the sampling points in the sorted sampling point data of the reflective columns meet a set distance range and the light intensity data of the sampling points meet a set light intensity range to obtain preprocessed data; dividing continuous data points of the sampling point angle data in the preprocessed data into a group to obtain a plurality of groups of data to be identified; and sequentially carrying out feature filtering and position extraction on each group of data to be identified to obtain a reflecting column position list. The method and the system for identifying the position of the reflecting column can accurately identify the actual position of the real reflecting column.

Description

Method and system for identifying position of reflecting column

Technical Field

The invention relates to the field of intelligent warehouse logistics, in particular to a method and a system for identifying the position of a reflecting column.

Background

Location and navigation technology based on 2D laser scanner are the key technique in fields such as industry AGV, intelligent robot, compare traditional rail navigation mode, and this technique has advantages such as positioning accuracy is high, nimble changeable, is applicable to in complicated, the high dynamic industry scene. The storage industry is the earliest place to apply the AGV, and at present, the AGV is widely applied to the intelligent storage logistics. Install rotatable 2D laser scanner on the AGV, install the laser positioning mark reflection of light post of high reflectivity on the wall or the pillar along the moving path, the AGV relies on 2D laser scanner transmission laser beam, and acquire 360 angle ranges's of horizontal plane reflected light intensity and distance data, vehicle-mounted computer discerns and obtains reflection of light post position and calculate the current position of vehicle and angle, compare the position of revising through the digital map with built-in, thereby realize automatic handling. However, the existing laser positioning technology based on the reflecting column at home and abroad is difficult to achieve high positioning precision in a complex industrial scene, and the practicability of the laser positioning and navigation technology is greatly restricted. One of the key factors that the existing laser positioning technology based on the reflective column is difficult to achieve high positioning accuracy is that the reflective column position identification process is poor in noise filtering performance, that is, environmental optical noise, such as glass mirror reflection noise, aluminum alloy reflection noise, and light reflection noise, cannot be filtered to a great extent, so that the actual position of the real reflective column cannot be accurately identified.

Disclosure of Invention

The invention aims to provide a method and a system for identifying the position of a reflective column, which can accurately identify the actual position of a real reflective column.

In order to achieve the purpose, the invention provides the following scheme:

a method for identifying the position of a reflective column comprises the following steps:

acquiring sampling point data of the reflective columns; the reflective column sampling point data comprises a plurality of groups of data points; each group of data points comprises sampling point light intensity data, sampling point distance data and sampling point angle data;

sorting the data points in the reflective column sampling point data according to the size of the sampling point angle data to obtain sorted reflective column sampling point data;

selecting data points, in the sorted light reflecting column sampling point data, of which the sampling point distance data meet a set distance range and the sampling point light intensity data meet a set light intensity range to obtain preprocessed data;

dividing continuous data points of the sampling point angle data in the preprocessed data into a group to obtain a plurality of groups of data to be identified;

and sequentially carrying out feature filtering and position extraction on the data to be identified in each group to obtain a position list of the reflection columns.

Optionally, the acquiring of the data of the sampling point of the reflective pillar specifically includes:

acquiring n groups of initial reflection column sampling point data scanned by a 2D laser scanner; each group of the initial reflection column sampling point data comprises sampling point light intensity data A_iAnd sample point distance data D_iTwo sampling values;

according to the formula ang_iCalculating light intensity data A of each group of sampling points at-180 + (360/n) × (n-1)_iAnd sample point distance data D_iCorresponding sampling point angle data ang_i；

Each group of the initial reflection column sampling point data and the sampling point angle data ang corresponding to the initial reflection column sampling point data_iAnd combining to obtain the sampling point data of the reflective column.

Optionally, the dividing the continuous data points of the sampling point angle data in the preprocessed data into a group to obtain multiple groups of data to be identified specifically includes:

determining continuous angle data of the sampling points in the preprocessed data;

and grouping the data points in the preprocessed data according to the continuous angle data of the sampling points in the preprocessed data to obtain multiple groups of data to be identified.

Optionally, in proper order to every group treat that the identification data carries out feature filtering and position extraction, obtain reflection of light post position list, specifically include:

sequentially judging whether the number of the data points in each group of the data to be identified is within a sampling distance threshold set value range, if so, retaining the whole group of the data to be identified, otherwise, giving up the whole group of the data to be identified, and obtaining a plurality of groups of data to be identified after primary characteristic filtering;

sequentially judging whether a first maximum value in each group of data to be identified after the first characteristic filtering is smaller than a first set threshold value, if so, retaining the whole group of data to be identified after the first characteristic filtering, and if not, abandoning the whole group of data to be identified after the first characteristic filtering to obtain a plurality of groups of data to be identified after the second characteristic filtering; the first maximum value is the maximum value in the absolute value of the difference of the light intensity data of every two adjacent sampling points in the data to be identified after the first characteristic filtering;

sequentially judging whether a second maximum value in each group of data to be identified after the second characteristic filtering is smaller than a second set threshold value, if so, retaining the whole group of data to be identified after the second characteristic filtering, and if not, abandoning the whole group of data to be identified after the second characteristic filtering to obtain a plurality of groups of data to be identified after the third characteristic filtering; the second maximum value is the maximum value in the absolute value of the difference between the distance data of every two adjacent sampling points in the data to be identified after the second characteristic filtering;

sequentially judging whether the error deviation percentage of the average value of all sampling point distance data in each group of the data to be identified after the third characteristic filtering and the theoretical distance is smaller than a third set threshold value or not, if so, retaining the whole group of the data to be identified after the third characteristic filtering, otherwise, giving up the whole group of the data to be identified after the third characteristic filtering, and obtaining a plurality of groups of data to be identified after the fourth characteristic filtering;

sequentially judging whether the average value of function values formed by light intensity data of every adjacent three sampling points in each group of data to be identified after the fourth characteristic filtering is larger than a fourth set threshold value, if so, retaining the whole group of data to be identified after the fourth characteristic filtering, and if not, abandoning the whole group of data to be identified after the fourth characteristic filtering to obtain a plurality of groups of data to be identified after the fifth characteristic filtering;

sequentially calculating the average value of angle data of all sampling points and the average value of distance data of all sampling points in the data to be identified after each group of fifth feature filtering to obtain a position list of the reflective columns; the light reflecting column position list comprises a plurality of position points, the number of the position points is the same as the group number of the data to be identified after the fifth feature filtering, and different position points correspond to different data to be identified after the fifth feature filtering; and each position point comprises the average value of angle data of all sampling points and the average value of distance data of all sampling points in the data to be identified after the fifth characteristic filtering.

Optionally, the sequentially determining whether the group number of the data points in each group of the data to be identified is within the range of the sampling distance threshold set value specifically includes:

according to the formula

And

calculating the range C of the set value of the sampling distance threshold_1h～C_1l(ii) a Wherein R is the radius of the column body of the reflective column, R_lAnd R_hRespectively setting a minimum distance and a maximum distance of the set distance range, wherein alpha is a sampling angle step length of the 2D laser scanner;

according to formula C_1h≤m≤C_1lSequentially judging whether the group number m of the data points in each group of the data to be identified is within a sampling distance threshold set value range C_1h～C_1lAnd (4) the following steps.

Optionally, the determining, in sequence, whether the error deviation percentage between the average value of the distance data of all sampling points in the data to be identified after each group of the third-time feature filtering and the theoretical distance is smaller than a third set threshold specifically includes:

step 1: according to the formula

Calculating the theoretical distance of the reflector when the group number of the data points in the data to be identified after the third time of characteristic filtering is p

In the formula, r is the radius of a reflecting column cylinder, and alpha is the sampling angle step length of the 2D laser scanner;

step 2: calculating the average value h of distance data of all sampling points in the data to be identified after the third characteristic filtering;

and step 3: according toFormula (II)

Judging the theoretical distance

Whether the error deviation percentage from the average value h is less than a third set threshold value C₄；

And 4, step 4: and repeating the steps 1-3 until the data to be identified after each group of the third-time feature filtration is judged.

Optionally, whether the average of the function values formed by the light intensity data of every adjacent three sampling points in the data to be identified after the fourth-time feature filtering in each group is larger than a fourth set threshold is sequentially judged, and the method specifically includes:

step 1: according to the formula z_i＝a_i+2-2*a_i+1+a_iCalculating light intensity data a of every adjacent three sampling points in the data to be identified after the fourth feature filtering_i+2Sampling point light intensity data a_i+1And light intensity data a of sampling point_iValue z of (A) of unevenness_i(ii) a The value of the concavity and convexity z_iRepresenting light intensity data a of every three adjacent sampling points_i+2Sampling point light intensity data a_i+1And light intensity data a of sampling point_iThe roughness of the formed curve, when the value of the roughness z is_iGreater than 0 is characterized by a convex curve when said value of concavity and convexity z is_iLess than or equal to 0 is characterized as a concave curve;

step 2: according to the formula

Calculating light intensity data a of every three adjacent sampling points_i+2Sampling point light intensity data a_i+1And light intensity data a of sampling point_iFunction value f (z) of composition_i)；

And step 3: according to the formula

Calculating the function value f (z)_i) Is determined as the average value of the function valuesf(z_i) Is greater than a fourth set threshold value C₅(ii) a Wherein q is the group number of the data points in the data to be identified after the fourth feature filtering;

and 4, step 4: and repeating the steps 1-3 until the data to be identified after each group of fourth feature filtering is judged.

Optionally, the average value of angle data of all sampling points and the average value of distance data of all sampling points in the data to be identified after each group of the fifth feature filtering are sequentially calculated to obtain a light reflecting column position list, which specifically includes:

step 1: according to the formula

Calculating angle data ang of all sampling points in the data to be identified after the fifth feature filtering_iAverage value of (A) O_ang(ii) a In the formula, s is the number of the sampling point angle data in the data to be identified after the fifth feature filtering;

step 2: according to the formula

Calculating all sampling point distance data d in the data to be identified after the fifth feature filtering_iAverage value of (A) O_d(ii) a In the formula, s is the number of the sampling point distance data in the data to be identified after the fifth feature filtering;

and step 3: according to the angle data ang of all sampling points in the data to be identified after the fifth feature filtering_iAverage value of (A) O_angAnd all sample point distance data d_iAverage value of (A) O_dObtaining the position points of the reflecting columns;

and 4, step 4: and repeating the steps 1-3 until the data to be identified after each group of fifth feature filtering all obtain position points, and all the position points jointly form a light reflecting column position list.

In order to achieve the above purpose, the invention also provides the following scheme:

a reflective post position identification system, comprising:

the acquisition module is used for acquiring sampling point data of the reflective column; the reflective column sampling point data comprises a plurality of groups of data points; each group of data points comprises sampling point light intensity data, sampling point distance data and sampling point angle data;

the sorting module is used for sorting the data points in the reflective column sampling point data according to the size of the sampling point angle data to obtain sorted reflective column sampling point data;

the preprocessing module is used for selecting data points, of the sorted light reflecting column sampling point data, of which the sampling point distance data meet a set distance range and the sampling point light intensity data meet a set light intensity range to obtain preprocessed data;

the grouping module is used for grouping continuous data points of the sampling point angle data in the preprocessed data to obtain a plurality of groups of data to be identified;

and the identification module is used for sequentially filtering the characteristics and extracting the positions of the data to be identified to obtain a reflecting column position list.

Optionally, the identification module specifically includes:

the first characteristic filtering unit is used for sequentially judging whether the group number of the data points in each group of the data to be identified is within a sampling distance threshold set value range, if so, retaining the whole group of the data to be identified, and if not, abandoning the whole group of the data to be identified to obtain a plurality of groups of data to be identified after the first characteristic filtering;

the second feature filtering unit is used for sequentially judging whether a first maximum value in each group of data to be identified after the first feature filtering is smaller than a first set threshold value, if so, retaining the whole group of data to be identified after the first feature filtering, and if not, abandoning the whole group of data to be identified after the first feature filtering to obtain a plurality of groups of data to be identified after the second feature filtering; the first maximum value is the maximum value in the absolute value of the difference of the light intensity data of every two adjacent sampling points in the data to be identified after the first characteristic filtering;

the third feature filtering unit is used for sequentially judging whether a second maximum value in each group of data to be identified after the second feature filtering is smaller than a second set threshold value, if so, retaining the whole group of data to be identified after the second feature filtering, and if not, abandoning the whole group of data to be identified after the second feature filtering to obtain a plurality of groups of data to be identified after the third feature filtering; the second maximum value is the maximum value in the absolute value of the difference between the distance data of every two adjacent sampling points in the data to be identified after the second characteristic filtering;

the fourth feature filtering unit is used for sequentially judging whether the error deviation percentage between the average value of all sampling point distance data in each group of the data to be identified after the third feature filtering and the theoretical distance is smaller than a third set threshold value, if so, retaining the whole group of the data to be identified after the third feature filtering, and if not, abandoning the whole group of the data to be identified after the third feature filtering to obtain a plurality of groups of data to be identified after the fourth feature filtering;

the fifth feature filtering unit is used for sequentially judging whether the average value of function values formed by light intensity data of every adjacent three sampling points in each group of the data to be identified after the fourth feature filtering is larger than a fourth set threshold value, if so, retaining the whole group of the data to be identified after the fourth feature filtering, and if not, abandoning the whole group of the data to be identified after the fourth feature filtering to obtain a plurality of groups of data to be identified after the fifth feature filtering;

the position extraction unit is used for sequentially calculating the average value of angle data of all sampling points and the average value of distance data of all sampling points in the data to be identified after each group of fifth feature filtering to obtain a reflecting column position list; the light reflecting column position list comprises a plurality of position points, the number of the position points is the same as the group number of the data to be identified after the fifth feature filtering, and different position points correspond to different data to be identified after the fifth feature filtering; and each position point comprises the average value of angle data of all sampling points and the average value of distance data of all sampling points in the data to be identified after the fifth characteristic filtering.

According to the specific embodiment provided by the invention, the invention discloses the following technical effects:

the invention discloses a method and a system for identifying the position of a reflecting column, which divide continuous data points of sampling point angle data into a group according to the continuity of the sampling point angle data, thereby realizing grouping the data points in the light reflecting column sampling point data to obtain a plurality of groups of data to be identified, sequentially carrying out feature filtration and position extraction on each group of data to be identified, because the data which meets the characteristic filtering condition is reserved and the data which does not meet the characteristic filtering condition is abandoned, the data after the characteristic filtering ensures that the environmental optical noise, including glass mirror reflection noise, aluminum alloy reflection noise, lamplight reflection noise and the like, is filtered to a greater extent, therefore, sampling data which accord with the characteristics of the reflective column after characteristic filtering can be obtained, the actual position of the real reflective column can be accurately identified according to the sampling data which accord with the characteristics of the reflective column, and an accurate reflective column position list can be obtained.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.

FIG. 1 is a flowchart illustrating a method for identifying a position of a reflective pillar according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of the data composition of the sampling points of the whole frame of the initial reflective columns;

FIG. 3 is a schematic diagram of a 2D laser scanner local coordinate system construction;

FIG. 4 is a schematic view of a reflective column;

FIG. 5 is a full frame of 2D laser scanner raw data saved in an experimental environment;

FIG. 6 is a schematic diagram of a graphical display of data of 1 group at around 0 °;

FIG. 7 is a flowchart illustrating feature filtering and location extraction in an embodiment of a method for identifying a position of a reflective prism according to the present invention;

FIG. 8 is a schematic diagram of a list of locations of reflective posts;

FIG. 9 is a diagram of a reflective column position identification system according to an embodiment of the present invention;

FIG. 10 is a diagram illustrating an identification module of an embodiment of a reflective prism position identification system according to the present invention;

FIG. 11 is a flowchart of the identification module operation.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below.

Fig. 1 is a flowchart illustrating a method for identifying a position of a reflective pillar according to an embodiment of the present invention. Referring to fig. 1, the method for identifying the position of the reflective pillar includes:

step 101: acquiring sampling point data of the reflective columns; the reflective column sampling point data comprises a plurality of groups of data points; each group of data points comprises sampling point light intensity data, sampling point distance data and sampling point angle data.

The step 101 specifically includes:

acquiring n groups of initial reflection column sampling point data scanned by a 2D laser scanner; the sampling angle range of the whole frame of data obtained by scanning of the 2D laser scanner is [ -180,180) ], fig. 2 is a schematic diagram of the whole frame of initial reflective column sampling point data composition, referring to fig. 2, each set of initial reflective column sampling point data comprises sampling point light intensity data A_iAnd sample point distance data D_iTwo are providedAnd (6) sampling values.

According to the formula ang_iCalculating light intensity data A of each group of sampling points at-180 + (360/n) × (n-1)_iAnd sample point distance data D_iCorresponding sampling point angle data ang_i。

In this embodiment, the 2D laser scanner is a 2D laser scanner widely used in industry, such as SICKNAV350 or becafur 2000, and its working principle is as follows: the 2D laser scanner takes mechanical 0 degrees as an initial angle, sequentially emits laser beams at various angles clockwise or anticlockwise by a certain sampling angle step length, collects the reflected light intensity of the laser beam angle, and calculates the distance of a reflector according to the emitting-receiving time difference, so that the distance of the reflector and the reflected light intensity at various angles within the range of 360 degrees are obtained.

FIG. 3 is a schematic diagram of the 2D laser scanner local coordinate system construction. Referring to fig. 3, a coordinate system is formed with a mechanical 0 ° of the 2D laser scanner as a positive x-axis direction and 90 ° thereof as a positive y-axis direction, and with a center of a sensor scanning circle thereof as an origin. Fig. 4 is a schematic view of a reflective column. According to the working principle of the 2D laser scanner, the reflective column usually uses a reflective sticker with high reflectivity to wrap the surface of the column, as shown in fig. 4. During the use, stand in 2D laser scanner's scanning plane with reflection of light cylinder to the scanner can detect out the position of reflection of light cylinder easily through the sudden change of light intensity. In this embodiment, the 2D laser scanner is set at the center positions of the plurality of reflective columns as shown in fig. 4, the sampling angle step length of the 2D laser scanner is set, the plurality of reflective columns are all located in the scanning plane of the 2D laser scanner, the 2D laser scanner performs annular scanning sampling with the sampling angle step length, sampling data is acquired, and the number of sampling points is set when the scanner starts initialization.

Step 102: and sorting the data points in the reflective column sampling point data according to the size of the sampling point angle data to obtain the sorted reflective column sampling point data.

Step 103: and selecting data points of which the sampling point distance data meet a set distance range and the sampling point light intensity data meet a set light intensity range from the sorted light reflecting column sampling point data to obtain preprocessed data.

In this embodiment, the original data obtained in step 101 is traversed in ascending or descending order of angle values, and data having a distance range between the set distance range Rl to Rh and a light intensity range between the set light intensity range Al to Ah are stored. Determining the length of the view radius of the 2D laser scanner as Rl-Rh according to the deployment condition of the reflective columns in the actual field, selecting the reflective column sampling point data in which the sampling point distance data in the sorted reflective column sampling point data meets the set distance range Rl-Rh, only processing the data of which the distance is greater than Rl and less than Rh, and ignoring the data outside the range to obtain preliminary pre-processing data. The set light intensity range is determined to be Al-Ah according to the reflected light intensity value range of the reflective column material in the view radius of the 2D laser scanner, and the reflected light intensity value range of the reflective column material in the view radius of the 2D laser scanner can be obtained through experimental comparison and statistics. And selecting the light intensity data of the sampling points in the preliminary preprocessing data to meet the data of the sampling points of the light intensity range Al-Ah, processing only the data with the distance greater than Al and less than Ah, and neglecting the data outside the range to obtain the preprocessing data.

Step 104: and dividing the continuous data points of the sampling point angle data in the preprocessed data into a group to obtain a plurality of groups of data to be identified.

The step 104 specifically includes:

and determining continuous angle data of the sampling points in the preprocessed data.

Fig. 5 is original data of the whole frame 2D laser scanner stored in the experimental environment, and is displayed visually by using a pythonmatplotlib module, where a horizontal axis in the graph is an angle, a vertical axis of an upper sub-graph is light intensity, and a vertical axis of a lower sub-graph is a distance. The sampling angle range of the whole frame of 2D laser scanner raw data is [ -180,180), it can be seen that the light intensity peaks at certain angles, which indicates that there may be a reflective column at that angle in the 2D laser scanner local coordinate system. If Rl is 0.3, Rh is 30, Al is 400, and Ah is 1800, 10 groups of data are obtained by grouping in fig. 5. The graphical display of the set 1 around 0 ° is shown in fig. 6.

Step 105: and sequentially carrying out feature filtering and position extraction on the data to be identified in each group to obtain a position list of the reflection columns.

Fig. 7 is a flowchart of feature filtering and position extracting in an embodiment of the method for identifying a position of a reflective column according to the present invention. Referring to fig. 7, the step 105 specifically includes:

step 701: and sequentially judging whether the group number of the data points in each group of the data to be identified is within the range of a sampling distance threshold set value, if so, retaining the whole group of the data to be identified, otherwise, abandoning the whole group of the data to be identified, and obtaining a plurality of groups of data to be identified after the first characteristic is filtered.

The step 701 specifically includes:

according to the formula

And

calculating the range C of the set value of the sampling distance threshold_1h～C_1l(ii) a Wherein R is the radius of the reflective column (i.e. the radius of the cylindrical reflective column), and R is_lAnd R_hAnd alpha is the sampling angle step length of the 2D laser scanner, and is respectively the minimum distance and the maximum distance (the closest visual field distance and the farthest visual field distance of the scanner) of the set distance range.

When the distance between the reflective columns is far, the data points of the reflective columns collected by the scanner are few, and when the distance between the reflective columns is near, the data points collected by the scanner are many, and the number of the data points is required to be within a certain range. m is the group number of the data points in the data to be identified in each group, namely the number of sampling points in the data to be identified in each group, when the number of the sampling points is within the range of a sampling distance threshold set value, the filtering requirement is met, otherwise, the group data identification is abandoned, whether all the data groups are identified is judged, and if no data group is identified, an unidentified group of data groups is selected for filtering identification.

Step 702: sequentially judging whether a first maximum value in each group of data to be identified after the first characteristic filtering is smaller than a first set threshold value, if so, retaining the whole group of data to be identified after the first characteristic filtering, and if not, abandoning the whole group of data to be identified after the first characteristic filtering to obtain a plurality of groups of data to be identified after the second characteristic filtering; the first maximum value is the maximum value in the absolute value of the difference between the light intensity data of every two adjacent sampling points in the data to be identified after the first characteristic filtering.

Data characteristics of light intensity deviation of adjacent sampling points:

and (3) judging standard:

max(|a_i+1-a_i|)<C₂

in the formula a_iAnd the light intensity data of the ith sampling point in the data to be identified after each group of the first characteristic filtering is obtained. C₂The threshold value is set for the first time, and is observed by a large amount of data.

Traversing absolute values of differences of all two adjacent light intensity data in the data to be identified after each group of the first characteristic filtering, selecting the maximum value, and judging whether the value is smaller than a first set threshold value C₂If it is less than C₂The filtering requirement is met, otherwise the identification of the group of data is abandoned. And judging whether all the data sets are identified or not, and if the data sets which are not identified exist, selecting an unidentified data set for filtering identification.

Step 703: sequentially judging whether a second maximum value in each group of data to be identified after the second characteristic filtering is smaller than a second set threshold value, if so, retaining the whole group of data to be identified after the second characteristic filtering, and if not, abandoning the whole group of data to be identified after the second characteristic filtering to obtain a plurality of groups of data to be identified after the third characteristic filtering; and the second maximum value is the maximum value in the absolute value of the difference between the distance data of every two adjacent sampling points in the data to be identified after the second characteristic filtering.

Data characteristics of adjacent sampling point distance deviation:

and (3) judging standard:

max(|d_i+1-d_i|)<C₃

in the formula d_iAnd the distance data of the ith sampling point in the data to be identified after the second characteristic filtering is carried out for each group. C₃The threshold is set for the second time, and is observed by a large amount of data.

Traversing the absolute value of the distance deviation between all two adjacent sampling points in the data to be identified after each group of second-time feature filtering, selecting the maximum value, and judging whether the maximum value is smaller than a second set threshold value C₃If it is less than C₃If not, abandoning the identification of the group of data, judging whether all data groups are identified, and if not, selecting an unidentified group of data groups for filtering identification.

Step 704: and sequentially judging whether the error deviation percentage of the average value of all sampling point distance data in the data to be identified after the third-time feature filtering and the theoretical distance in each group is smaller than a third set threshold value, if so, retaining the whole group of the data to be identified after the third-time feature filtering, and if not, giving up the whole group of the data to be identified after the third-time feature filtering to obtain a plurality of groups of data to be identified after the fourth-time feature filtering.

This step 704 specifically includes:

step 1: according to the formula

Calculating the group number of the data points in the data to be identified after the third time characteristic filtering to be pTheoretical distance of reflecting column

In the formula, r is the radius of the reflecting column cylinder, and alpha is the sampling angle step length of the 2D laser scanner.

Step 2: and calculating the average value h of distance data of all sampling points in the data to be identified after the third time of characteristic filtering.

And step 3: according to the formula

Judging the theoretical distance

Whether the error deviation percentage from the average value h is less than a third set threshold value C₄。

p is the group number of the data points in the data to be identified after the third time of characteristic filtering, namely the number of sampling points in the data to be identified after the third time of characteristic filtering in each group, theoretically, the number of the sampling points and the sampling distance are in a linear relation, so that the theoretical distance of the reflective column can be calculated according to the number of the sampling points, the error deviation percentage of the theoretical distance and the actual sampling distance is calculated, and whether the value is smaller than a third set threshold value C or not is judged₄Third setting threshold C₄Observed from a large amount of data, if the data is less than a third set threshold value C₄If not, abandoning the identification of the group of data, judging whether all data groups are identified, and if not, selecting an unidentified group of data groups for filtering identification.

Step 705: and sequentially judging whether the average value of function values formed by light intensity data of every adjacent three sampling points in the data to be identified after the fourth characteristic filtering is larger than a fourth set threshold value or not, if so, retaining the whole group of the data to be identified after the fourth characteristic filtering, otherwise, abandoning the whole group of the data to be identified after the fourth characteristic filtering, and obtaining a plurality of groups of data to be identified after the fifth characteristic filtering.

The step 705 specifically includes:

step 1: according to the formula z_i＝a_i+2-2*a_i+1+a_iCalculating light intensity data a of every adjacent three sampling points in the data to be identified after the fourth feature filtering_i+2Sampling point light intensity data a_i+1And light intensity data a of sampling point_iValue z of (A) of unevenness_i(ii) a The value of the concavity and convexity z_iRepresenting light intensity data a of every three adjacent sampling points_i+2Sampling point light intensity data a_i+1And light intensity data a of sampling point_iThe roughness of the formed curve, when the value of the roughness z is_iGreater than 0 is characterized by a convex curve when said value of concavity and convexity z is_i0 or less is characterized as a concave curve.

Step 2: according to the formula

Calculating light intensity data a of every three adjacent sampling points_i+2Sampling point light intensity data a_i+1And light intensity data a of sampling point_iFunction value f (z) of composition_i)。

And step 3: according to the formula

Calculating the function value f (z)_i) Is determined as the average value of the function values f (z)_i) Is greater than a fourth set threshold value C₅(ii) a Wherein q is the number of groups of the data points in the data to be identified after the fourth feature filtering.

q is the number of groups of the data points in the data to be identified after the fourth feature filtering, namely the number of sampling points in the data to be identified after the fourth feature filtering in each group, z_i＝a_i+2-2*a_i+1+a_iRepresenting the light intensity curve of each group of dataRelief of 3 successive points in the line, when z_i>0 indicates that it is a convex curve, whereas it is a concave curve. All z's for each set of data are computed sequentially_i,i∈[0,q-2]And the values are accumulated and then averaged to judge whether the value is less than a fourth set threshold value C₅Fourth, a threshold value C is set₅Observed from a large amount of data, if the value is less than a fourth set threshold value C₅If not, abandoning the identification of the group of data, judging whether all data groups are identified, and if not, selecting an unidentified group of data groups for filtering identification.

Step 706: sequentially calculating the average value of angle data of all sampling points and the average value of distance data of all sampling points in the data to be identified after each group of fifth feature filtering to obtain a position list of the reflective columns; the light reflecting column position list comprises a plurality of position points, the number of the position points is the same as the group number of the data to be identified after the fifth feature filtering, and different position points correspond to different data to be identified after the fifth feature filtering; and each position point comprises the average value of angle data of all sampling points and the average value of distance data of all sampling points in the data to be identified after the fifth characteristic filtering.

The step 706 specifically includes:

step 1: according to the formula

step 2: according to the formula

Calculating all sampling point distance data d in the data to be identified after the fifth feature filtering_iAverage value of (A) O_d(ii) a In the formula, sThe number of the sampling point distance data in the data to be identified after the fifth feature filtering is obtained;

Fig. 8 is a schematic diagram of a light-reflecting column position list, and referring to fig. 8, each group of the data to be identified after the fifth feature filtering obtains a group of position points including angle data and distance data.

The method for identifying the position of the reflecting column has extremely high noise filtering performance, and can filter more than 98% of environmental optical noise, such as glass mirror reflection noise, aluminum alloy reflection noise, lamplight reflection noise and the like, to obtain sampling data conforming to the characteristics of the reflecting column, so that the real reflecting column is better identified, and the accuracy of identifying the real reflecting column is obviously improved.

Fig. 9 is a structural diagram of a reflective column position identification system according to an embodiment of the invention. Referring to fig. 9, the reflective column position identification system includes:

an obtaining module 901, configured to obtain data of sampling points of the reflective columns; the reflective column sampling point data comprises a plurality of groups of data points; each group of data points comprises sampling point light intensity data, sampling point distance data and sampling point angle data.

The obtaining module 901 specifically includes:

the device comprises an initial reflective column sampling point data acquisition unit, a data acquisition unit and a data acquisition unit, wherein the initial reflective column sampling point data acquisition unit is used for acquiring n groups of initial reflective column sampling point data scanned by a 2D laser scanner; each group of the initial reflection column sampling point data comprises sampling point light intensity data A_iAnd sample point distance data D_iTwo sampled values.

A sampling point angle data calculating unit for calculating the angle of the sampling point according to the formula ang_i-180+ (360/n) × (n-1) counts perGroup sampling point light intensity data A_iAnd sample point distance data D_iCorresponding sampling point angle data ang_i。

A reflective column sampling point data acquisition unit for acquiring each set of the initial reflective column sampling point data and the sampling point angle data ang corresponding to the initial reflective column sampling point data_iAnd combining to obtain the sampling point data of the reflective column.

And the sorting module 902 is configured to sort the data points in the reflective pillar sampling point data according to the size of the sampling point angle data, so as to obtain sorted reflective pillar sampling point data.

The preprocessing module 903 is configured to select data points, of the sorted light reflecting column sampling point data, where the sampling point distance data meets a set distance range and the sampling point light intensity data meets a set light intensity range, to obtain preprocessed data.

And a grouping module 904, configured to group data points of the preprocessed data, where the angle data of the sampling points is continuous, into a group, so as to obtain multiple groups of data to be identified.

The grouping module 904 specifically includes:

and the continuous sampling point angle data determining unit is used for determining continuous sampling point angle data in the preprocessed data.

And the grouping unit is used for grouping the data points in the preprocessed data according to the continuous angle data of the sampling points in the preprocessed data to obtain multiple groups of data to be identified.

And the identification module 905 is used for sequentially filtering and extracting the characteristics and the positions of the data to be identified to obtain a reflecting column position list.

FIG. 10 is a block diagram of an identification module in an embodiment of a reflective prism position identification system according to the present invention. Referring to fig. 10, the identification module 905 specifically includes:

the first feature filtering unit 1001 is configured to sequentially determine whether the number of groups of data points in each group of data to be identified is within a range of a sampling distance threshold setting value, if so, keep the entire group of data to be identified, otherwise, abandon the entire group of data to be identified, and obtain multiple groups of data to be identified after the first feature filtering.

The first feature filter unit 1001 specifically includes:

a sub-unit for calculating the range of the set value of the sampling distance threshold according to a formula

And

calculating the range C of the set value of the sampling distance threshold_1h～C_1l(ii) a Wherein R is the radius of the column body of the reflective column, R_lAnd R_hAnd respectively the minimum distance and the maximum distance of the set distance range, wherein alpha is the sampling angle step length of the 2D laser scanner.

A first judging subunit for judging according to formula C_1h≤m≤C_1lSequentially judging whether the group number m of the data points in each group of the data to be identified is within a sampling distance threshold set value range C_1h～C_1lAnd (4) the following steps.

The second feature filtering unit 1002 is configured to sequentially determine whether a first maximum value in each group of data to be identified after the first feature filtering is smaller than a first set threshold, if so, keep the entire group of data to be identified after the first feature filtering, otherwise, abandon the entire group of data to be identified after the first feature filtering, and obtain multiple groups of data to be identified after the second feature filtering; the first maximum value is the maximum value in the absolute value of the difference between the light intensity data of every two adjacent sampling points in the data to be identified after the first characteristic filtering.

A third feature filtering unit 1003, configured to sequentially determine whether a second maximum value in each group of data to be identified after the second feature filtering is smaller than a second set threshold, if so, retain the entire group of data to be identified after the second feature filtering, otherwise, abandon the entire group of data to be identified after the second feature filtering, and obtain multiple groups of data to be identified after the third feature filtering; and the second maximum value is the maximum value in the absolute value of the difference between the distance data of every two adjacent sampling points in the data to be identified after the second characteristic filtering.

And the fourth feature filtering unit 1004 is configured to sequentially determine whether error deviation percentages of average values of distance data of all sampling points in the data to be identified after each group of the third feature filtering and the theoretical distances are smaller than a third set threshold, if so, keep the whole group of the data to be identified after the third feature filtering, and if not, abandon the whole group of the data to be identified after the third feature filtering, so as to obtain multiple groups of data to be identified after the fourth feature filtering.

The fourth characteristic filtering unit 1004 specifically includes:

a theoretical distance calculating subunit of the reflecting column for calculating the theoretical distance according to a formula

And the sampling point distance average value operator unit is used for calculating the average value h of all sampling point distance data in the data to be identified after the third time of characteristic filtering.

A second judgment subunit for judging whether the first judgment subunit is a formula

Judging the theoretical distance

A fifth feature filtering unit 1005, configured to sequentially determine whether an average value of function values formed by light intensity data of every adjacent three sampling points in each set of data to be identified after the fourth feature filtering is greater than a fourth set threshold, if so, retain the whole set of data to be identified after the fourth feature filtering, otherwise, abandon the whole set of data to be identified after the fourth feature filtering, and obtain multiple sets of data to be identified after the fifth feature filtering.

The fifth characteristic filtering unit 1005 specifically includes:

a concave-convex value calculating operator unit for calculating the value of the convex-concave value according to the formula z_i＝a_i+2-2*a_i+1+a_iCalculating light intensity data a of every adjacent three sampling points in the data to be identified after the fourth feature filtering_i+2Sampling point light intensity data a_i+1And light intensity data a of sampling point_iValue z of (A) of unevenness_i(ii) a The value of the concavity and convexity z_iRepresenting light intensity data a of every three adjacent sampling points_i+2Sampling point light intensity data a_i+1And light intensity data a of sampling point_iThe roughness of the formed curve, when the value of the roughness z is_iGreater than 0 is characterized by a convex curve when said value of concavity and convexity z is_i0 or less is characterized as a concave curve.

A function value calculating operator unit for calculating a function value according to a formula

A third judging subunit for judging according to the formula

The position extraction unit 1006 is configured to sequentially calculate an average value of angle data of all sampling points and an average value of distance data of all sampling points in the data to be identified after each group of fifth feature filtering is performed, so as to obtain a light reflection column position list; the light reflecting column position list comprises a plurality of position points, the number of the position points is the same as the group number of the data to be identified after the fifth feature filtering, and different position points correspond to different data to be identified after the fifth feature filtering; and each position point comprises the average value of angle data of all sampling points and the average value of distance data of all sampling points in the data to be identified after the fifth characteristic filtering.

The position extracting unit 1006 specifically includes:

an operator unit for average value of sampling point angle according to formula

Calculating angle data ang of all sampling points in the data to be identified after the fifth feature filtering_iAverage value of (A) O_ang(ii) a And in the formula, s is the number of the sampling point angle data in the data to be identified after the fifth feature filtering.

An operator unit for average value of distance between sampling points according to a formula

Calculating all sampling point distance data d in the data to be identified after the fifth feature filtering_iAverage value of (A) O_d(ii) a Wherein s is the number of the sampling point distance data in the data to be identified after the fifth feature filtering.

A reflection column position point obtaining subunit, configured to obtain angle data ang of all sampling points in the data to be identified filtered according to the fifth feature_iAverage value of (A) O_angAnd all sample point distance data d_iAverage value of (A) O_dAnd obtaining the position points of the reflecting columns.

And obtaining position points of the data to be identified after each group of fifth feature filtering, wherein all the position points jointly form a light reflecting column position list. The reflecting column position list is the final output of the identification module.

Each feature filter unit in the identification module describes a data feature, which is a unique data feature of the reflection column data obtained by observing a large amount of data similar to the data of FIG. 6. The data characteristic fluctuates in a certain numerical range [ A-B ], when the data characteristic of the data in the data packet to be identified is extracted and the numerical value is within [ A-B ], the data passes through the characteristic filtering unit, and if the numerical value is out of [ A-B ], the data does not pass through the characteristic filtering unit.

The design steps are as follows:

1. some feature is inductively extracted from a large amount of actual sampled data.

2. A matching criterion for the feature is set.

The working process is as follows:

1. extracting the features of the group of data for the feature filter unit.

2. And judging whether the matching standard of the characteristic filtering unit is met.

Fig. 11 is a flow chart of the identification module, see fig. 11, the identification module is composed of 5 feature filtering units (feature filters) and 1 position extracting unit in serial link, wherein the feature filtering units 1 to 5 have data feature 1 to 5 extracting functions, and are provided with pass indexes, that is, the first feature filtering unit, the second feature filtering unit, the third feature filtering unit, the fourth feature filtering unit, and the fifth feature filtering unit have data features of sampling point number and sampling distance threshold setting value, data features of adjacent sampling point light intensity deviation, data features of adjacent sampling point distance deviation, data features of sampling distance and sampling point number, and data feature extracting functions of light intensity sampling values, and are provided with pass indexes. When the identification module works, the data to be identified sequentially pass through each feature filtering unit in a group order for identification, when the data passes through a certain feature filtering unit, firstly extracting the features of the data packet to be identified aiming at the filtering unit, judging whether the data meets the matching standard of the feature filtering unit, and if the data meets the requirement of the feature filtering unit, continuing to push backwards through the feature filtering unit; otherwise, if the data packet does not meet the requirement of the feature filtering unit, the feature filtering unit outputs no data packet, abandons the identification of the data packet to be identified and the data packet to be identified, judges whether all the data packets are identified, and sends the next data packet to be identified to the identification module and sequentially identifies the feature filtering unit if the data packets which are not identified exist.

The reflecting column position identification system disclosed by the invention has the advantages that the sampling data are subjected to characteristic filtration through each characteristic filtration unit to obtain the sampling data which accord with the characteristics of the reflecting column, and the sampling data which accord with the characteristics of the reflecting column are calculated through the position extraction unit to obtain the high-precision real reflecting column position data, so that the identification precision of the real reflecting column is obviously improved, the system stability is strong, the structure is simple, and the implementation is easy.

The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the system part for description.

The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.

Claims

1. A method for identifying the position of a reflective column is characterized by comprising the following steps:

sequentially carrying out feature filtering and position extraction on each group of data to be identified to obtain a position list of the reflective columns;

wherein, in proper order to every group treat that the identification data carries out feature filtering and position extraction, obtain reflection of light post position list, specifically include:

sequentially calculating the average value of angle data of all sampling points and the average value of distance data of all sampling points in the data to be identified after each group of fifth feature filtering to obtain a position list of the reflective columns; the light reflecting column position list comprises a plurality of position points, the number of the position points is the same as the group number of the data to be identified after the fifth feature filtering, and different position points correspond to different data to be identified after the fifth feature filtering; each position point comprises an average value of angle data of all sampling points and an average value of distance data of all sampling points in the data to be identified after the fifth characteristic filtering;

wherein, judge every group in proper order treat after the fourth feature filters whether the average value of the function value that every adjacent three sampling point light intensity data constitutes is greater than the fourth and sets for the threshold value in the data of waiting to discern, specifically include:

step 1: according to the formula z_i＝a_i+2-2*a_i+1+a_iCalculating the fourth time characteristic filtered data to be identified, wherein the light intensity data a of every three adjacent sampling points_i+2Sampling point light intensity data a_i+1And light intensity data a of sampling point_iValue z of (A) of unevenness_i(ii) a The value of the concavity and convexity z_iRepresenting light intensity data a of every three adjacent sampling points_i+2Sampling point light intensity data a_i+1And light intensity data a of sampling point_iThe roughness of the formed curve, when the value of the roughness z is_iGreater than 0 is characterized by a convex curveWhen the value of the concavity and convexity z is_iLess than or equal to 0 is characterized as a concave curve;

step 2: according to the formula

And step 3: according to the formula

Calculating the function value f (z)_i) Is determined as the average value of the function values f (z)_i) Is greater than a fourth set threshold value C₅(ii) a Wherein q is the group number of the data points in the data to be identified after the fourth feature filtering;

and 4, step 4: repeating the steps 1-3 until the data to be identified after each group of fourth feature filtering is judged;

q is the number of groups of the data points in the data to be identified after the fourth feature filtering, namely the number of sampling points in the data to be identified after the fourth feature filtering in each group, z_i＝a_i+2-2*a_i+1+a_iRepresenting the concave-convex of continuous 3 points in each group of data light intensity curve when z is_iIf the curve is more than 0, the curve is a convex curve, otherwise, the curve is a concave curve; all z's for each set of data are computed sequentially_i,i∈[0,q-2]And the values are accumulated and then averaged to judge whether the value is less than a fourth set threshold value C₅Fourth, a threshold value C is set₅Observed from the data, if less than the fourth set threshold C₅If not, abandoning the identification of the group of data, judging whether all data groups are identified, and if not, selecting an unidentified group of data groups for filtering identification.

2. The method for identifying the position of the reflective column according to claim 1, wherein the acquiring of the data of the sampling point of the reflective column specifically comprises:

according to the formula ang_iCalculating light intensity data A of each group of sampling points at-180 + (360/n) × (n-1)_iAnd sample point distance data D_iCorresponding sampling point angle data ang_i(ii) a n represents the numerical value of the group number of each group of initial reflection column sampling point data;

3. The method for identifying the position of the reflective column according to claim 1, wherein the step of dividing the continuous data points of the angle data of the sampling points in the preprocessed data into a group to obtain a plurality of groups of data to be identified specifically comprises:

4. The method for identifying a position of a reflective column according to claim 1, wherein the sequentially determining whether the number of groups of the data points in each group of the data to be identified is within a range of a set value of a sampling distance threshold comprises:

according to the formula

And

5. The method for identifying the position of the reflective column according to claim 1, wherein the sequentially determining whether the error deviation percentage between the average value of the distance data of all sampling points in the data to be identified after each group of the third-time feature filtering and the theoretical distance is smaller than a third set threshold specifically comprises:

step 1: according to the formula

and step 3: according to the formula

Judging the theoretical distance

6. The method for identifying the position of the reflective column according to claim 1, wherein the step of sequentially calculating an average value of angle data of all sampling points and an average value of distance data of all sampling points in the data to be identified after each group of the fifth feature filtering step to obtain a reflective column position list specifically comprises:

step 1: according to the formula

step 2: according to the formula

7. A reflective post position identification system, comprising:

the identification module is used for sequentially carrying out feature filtering and position extraction on each group of data to be identified to obtain a position list of the reflective columns;

wherein, the identification module specifically includes:

the position extraction unit is used for sequentially calculating the average value of angle data of all sampling points and the average value of distance data of all sampling points in the data to be identified after each group of fifth feature filtering to obtain a reflecting column position list; the light reflecting column position list comprises a plurality of position points, the number of the position points is the same as the group number of the data to be identified after the fifth feature filtering, and different position points correspond to different data to be identified after the fifth feature filtering; each position point comprises an average value of angle data of all sampling points and an average value of distance data of all sampling points in the data to be identified after the fifth characteristic filtering;

step 1: according to the formula z_i＝a_i+2-2*a_i+1+a_iCalculating the fourth time characteristic filtered data to be identified, wherein the light intensity data a of every three adjacent sampling points_i+2Sampling point light intensity data a_i+1And light intensity data a of sampling point_iValue z of (A) of unevenness_i(ii) a The value of the concavity and convexity z_iRepresenting light intensity data a of every three adjacent sampling points_i+2Sampling point light intensity data a_i+1And light intensity data a of sampling point_iThe roughness of the formed curve, when the value of the roughness z is_iGreater than 0 is characterized by a convex curve when said value of concavity and convexity z is_iLess than or equal to 0 is characterized as a concave curve;

step 2: according to the formula

And step 3: according to the formula

q is the number of groups of the data points in the data to be identified after the fourth feature filtering, namely the number of sampling points in the data to be identified after the fourth feature filtering in each group, z_i＝a_i+2-2*a_i+1+a_iRepresenting the concave-convex of continuous 3 points in each group of data light intensity curve when z is_iGreater than 0 indicates that it is a convex curve, andit is a concave curve; all z's for each set of data are computed sequentially_i,i∈[0,q-2]And the values are accumulated and then averaged to judge whether the value is less than a fourth set threshold value C₅Fourth, a threshold value C is set₅Observed from the data, if less than the fourth set threshold C₅If not, abandoning the identification of the group of data, judging whether all data groups are identified, and if not, selecting an unidentified group of data groups for filtering identification.