CN109910881B - Method and system for detecting and processing visual blind spot information of turning road - Google Patents

Abstract

The invention discloses a method and system for detecting and processing visual blind area information on a turning road. The method firstly detects the objects traveling on the turning road, performs clutter removal processing, and performs obstacle detection. The sensor data and the microwave sensor data are fused; the fused data is compared with the hexadecimal feature vector samples, and the microcontroller sends the detection data to the LED display for display and activates the warning light for warning. The detector sends the detection data and location information to the background server for classification and storage, and the background server performs fuzzy cluster analysis on the accumulated data to obtain a more accurate feature vector interval, and periodically sends the feature vector interval as the updated feature vector sample set to the microcontroller. The present invention can timely inform the driver of the visual blind spot information of the turning road in advance, thereby reducing the occurrence rate of traffic accidents.

Description

Method and system for detecting and processing visual blind spot information of turning road

technical field

The invention relates to a method and system for detecting and processing visual blind area information of a turning road, belonging to the technical field of road traffic information detection.

Background technique

In real life, there are often obstacles on the inside of the turning road to block the line of sight, forming a blind spot of vision. Therefore, the corner of the road is a frequent traffic accident area.

At present, there are many methods of road traffic detection, and they are becoming more and more mature, but most of them are for traffic flow and accident monitoring on urban conventional roads. For turning roads, currently, apart from widely used road wide-angle mirrors and static turn warning signs, there is almost no efficient intelligent detection and processing method, which is to accurately process the traffic information of turning roads and accurately identify moving objects on the road. Category, such as: pedestrian, vehicle category.

In order to inform the driver of the information in advance and reduce the incidence of traffic accidents, it is of great practical significance to study an information detection and processing method for the visual blind spot of the turning road.

SUMMARY OF THE INVENTION

In view of the shortcomings of the above methods, the present invention proposes a method and system for detecting and processing visual blind spot information on a turning road, which can timely inform the driver of the visual blind spot information on a turning road and reduce the incidence of traffic accidents.

The technical scheme adopted by the present invention to solve its technical problems is:

A method for detecting and processing visual blind spot information of a turning road provided by the present invention includes the following steps:

Step 1, use the geomagnetic sensor to detect the driving object on the turning road, if the vehicle is detected, send the magnetic field data of the Z-axis direction of the geomagnetic sensor to the microcontroller, otherwise go to step 3;

Step 2, the microcontroller carries out clutter removal processing to the magnetic field data of the Z-axis direction of the geomagnetic sensor, and stores the effective geomagnetic sensor data information;

Step 3, use the microwave sensor to detect the obstacle, and send the received microwave signal reflected by the obstacle to the microcontroller;

Step 4, the microcontroller stores the received microwave sensor data information, and performs data fusion with the received geomagnetic sensor data and microwave sensor data;

Step 5, the microcontroller compares the fused data with the hexadecimal feature vector sample, and if the matching is successful, then carry out the measurement and display of step 6, otherwise return to step 1;

Step 6, the radar sensor sends the detected data to the microcontroller, the microcontroller sends the detected data to the LED display screen for display and starts the warning light for warning, and the microcontroller sends the detected data and location information to the background server at the same time;

Step 7: The background server classifies and stores the received data, and performs fuzzy cluster analysis on the accumulated data to obtain a more accurate feature vector interval, and periodically sends the feature vector interval as an updated feature vector sample set to all the data. described microcontroller.

As a possible implementation of this embodiment, in step 2, the process of performing clutter removal processing on the magnetic field data in the Z-axis direction of the geomagnetic sensor is: reading the magnetic field data in the Z-axis direction of the geomagnetic sensor, The incoming data at the moment is used as a reference, and the Kalman filter algorithm is used to compare the incoming data at the present moment, infer the wrong data in the moment data, remove the wrong data caused by irrelevant factors, and extract useful data information.

As a possible implementation manner of this embodiment, in step 4, the specific process of data fusion is as follows: storing the data transmitted by the geomagnetic sensor in the first seven bits of an eight-bit binary number pre-created by the single-chip microcomputer; The sensor data is put into the eighth bit, and then the eight-bit binary data is converted into hexadecimal data.

As a possible implementation of this embodiment, in step 5, the specific process of comparing the fused data with the hexadecimal feature vector samples is as follows: the fused data is compared with the hexadecimal feature vector samples. Loop detection, the fused data is matched with each feature vector sample, and if the matching is successful, the information corresponding to the feature vector sample is returned.

As a possible implementation of this embodiment, in step 6, the microcontroller sends the detection data to the LED display screen for display and activates the warning light for warning. The specific process is: According to the incoming data, the speed of the moving object and the distance information of the moving object from the corner are calculated, and according to the corresponding type of the moving object, different data is called and sent to the LED display for display and the warning light is activated to warn.

As a possible implementation mode of the present embodiment, in step 7, the specific process that the background server performs fuzzy cluster analysis on the accumulated data is:

Step 71, the background server takes the latest n sets of fused geomagnetic data and microwave data from the database as a set of initial sample universe, and set each set of data as X ₁ , X ₂ ,...,X _m , and each data has m Each index represents its characteristics, then the characteristic attribute data of the i-th classification object in the initial sample universe is X _i1 ',X _i2 ',...,X _im ';

Step 72, using the following formula to standardize the feature attribute data:

in,

取适当的M值，使r_ij在[0,1]中 分散开；Step 73: Establish a fuzzy similarity relationship matrix, and determine the similarity matrix R=(r _ij ) _n×n by common cluster analysis, where r _ij is the similarity coefficient between X _i and X _j , where

Take an appropriate value of M to make r _ij spread out in [0,1];

Step 74: Perform a transitive closure operation on the similarity matrix R to obtain t(R)=R ⁿ , and on this basis, set the confidence level λ to take different values to obtain different clustering results.

The invention provides a detection and processing system for visual blind area information of a turning road, which includes a geomagnetic sensor, a microwave sensor, a radar sensor, a microcontroller, a communication module, an LED display screen and a warning light, which are arranged at the scene of the turning road. A background server for remote communication with a microcontroller, the geomagnetic sensor, microwave sensor, radar sensor, communication module, LED display screen and warning light are respectively connected to the microcontroller, and the geomagnetic sensor and the microwave sensor are set at the corner of the turning road On the forward straight lane, and the geomagnetic sensor and the microwave sensor are arranged in front and behind in the driving direction of the straight lane, the radar sensor is arranged at the corner of the turning road, and the probe of the radar sensor faces the straight lane in front of the corner of the turning road. The microcontroller, the communication module, the LED display and the warning light are arranged at the corner of the turning road, and the LED display faces the straight lane ahead of the corner of the turning road.

As a possible implementation of this embodiment, the geomagnetic sensor adopts the HMC5883L three-axis geomagnetic sensor, which is used to output the detected moving object data by detecting the metal content of the moving object; the microwave sensor adopts the model RCWL-0516 The microwave sensor is used for whether there is a moving object; the radar sensor adopts the TF03 lidar sensor, which is used to detect the speed of the moving object and the distance information to the turning angle.

As a possible implementation manner of this embodiment, the communication module adopts a SIM868 GSM/GPRS/GPS module.

As a possible implementation of this embodiment, the detection and processing system is applied to a one-way lane road and/or a two-way lane road.

The beneficial effects that the technical solutions of the embodiments of the present invention can have are as follows:

A method for detecting and processing visual blind spot information on a turning road according to an embodiment of the present invention detects objects traveling on the turning road, performs clutter removal processing, and performs obstacle detection. The data is fused; the fused data is compared with the hexadecimal feature vector samples, the microcontroller sends the detection data to the LED display for display and starts the warning light for warning, and the microcontroller compares the detection data with the hexadecimal feature vector samples. The location information is sent to the background server for classification and storage, and the background server performs fuzzy cluster analysis on the accumulated data to obtain a more accurate feature vector interval, and periodically sends the feature vector interval as an updated feature vector sample set to the micro-controller device. The technical solution of the embodiment of the present invention can timely inform the driver of the information of the visual blind spot of the turning road in advance, thereby reducing the occurrence rate of traffic accidents.

The technical solution of the embodiment of the present invention uses two sensors to output data, realizes the electronization of output information, and is more accurate than the traditional wide-angle lens in which the opposite road information is visually detected by the human eye; at the same time, the Kalman filter algorithm is used to The abnormal data on the road is effectively filtered out; the fusion algorithm is adopted to fuse the data from multiple sensors, which is convenient for data transmission, and only one data transmission can send all the data to improve the transmission efficiency. Different from the traditional threshold detection method, the fuzzy clustering analysis method is used to more accurately and efficiently identify the type information of moving objects, such as pedestrians, cyclists, and different types of vehicles, and accurately mine useful data. This detection method can be used in various places such as urban detours, mountain detours, residential quarters, underground parking lots, etc., and has a wide range of applications.

The system for detecting and processing visual blind spot information on a turning road according to the embodiment of the present invention can achieve the same beneficial effects as the method for detecting and processing visual blind spot information on a turning road according to the embodiment of the present invention.

Description of drawings:

Fig. 1 is a flow chart of a method for detecting and processing visual blind spot information of a turning road shown according to an exemplary embodiment;

Fig. 2 is a schematic block diagram of a system for detecting and processing visual blind spot information of a turning road according to an exemplary embodiment;

Fig. 3 is the front view of the combination of uprights on the side A side of the turning angle deviating road and the side B side of the turning angle deviating road in the specific implementation of the detection and processing system of the present invention, in the figure, 1-upright column, 2-LED screen, 3-warning light, 4-power supply, 5- -The sealed box includes: microcontroller, SIM868GSM/GPRS/GPS module;

Fig. 4 is the front view of the column combination at the corner in the specific implementation of the detection processing system of the present invention, in the figure, 6-radar sensor, 7-solar panel;

5 is a top view of the hardware location layout of the detection and processing system of the present invention, in the figure, 8- geomagnetic sensor A, 9- microwave sensor A, 10- corner deviating road A side column and microcontroller, display screen, warning light combination (specifically As shown in Figure 3), 11- Corner deviation road B side column, solar panel, radar sensor, 12- column, microcontroller, display screen, warning light combination (specifically shown in Figure 3), 13- Geomagnetic sensor B , 14-Microwave sensor B;

Fig. 6 is a specific flow chart of the present invention for detecting and processing the visual blind spot information of the turning road.

Detailed ways

Below in conjunction with accompanying drawing and embodiment, the present invention will be further described:

In order to clearly illustrate the technical features of this solution, the present invention will be described in detail below through specific embodiments and in conjunction with the accompanying drawings. The following disclosure provides many different embodiments or examples for implementing different structures of the invention. In order to simplify the disclosure of the present invention, the components and arrangements of specific examples are described below. Furthermore, the present invention may repeat reference numerals and/or letters in different instances. This repetition is for the purpose of simplicity and clarity and does not in itself indicate a relationship between the various embodiments and/or arrangements discussed. It should be noted that the components illustrated in the figures are not necessarily drawn to scale. Descriptions of well-known components and processing techniques and processes are omitted from the present invention to avoid unnecessarily limiting the present invention.

In order to clearly and intuitively illustrate the effect achieved by the present invention, the present invention will be described in detail below by taking the main urban area of Beijing as an example, through specific embodiments and in conjunction with the accompanying drawings.

Fig. 1 is a flowchart of a method for detecting and processing visual blind spot information of a turning road according to an exemplary embodiment. As shown in Figure 1, a method for detecting and processing visual blind spot information of a turning road provided by an embodiment of the present invention includes the following steps:

Step 1, use the geomagnetic sensor to detect the driving object on the turning road, if the vehicle is detected, send the magnetic field data of the Z-axis direction of the geomagnetic sensor to the microcontroller, otherwise go to step 3;

Step 2, the microcontroller carries out clutter removal processing to the magnetic field data of the Z-axis direction of the geomagnetic sensor, and stores the effective geomagnetic sensor data information;

Step 3, use the microwave sensor to detect the obstacle, and send the received microwave signal reflected by the obstacle to the microcontroller;

Step 4, the microcontroller stores the received microwave sensor data information, and performs data fusion with the received geomagnetic sensor data and microwave sensor data;

Step 5, the microcontroller compares the fused data with the hexadecimal feature vector sample, and if the matching is successful, then carry out the measurement and display of step 6, otherwise return to step 1;

Step 6, the radar sensor sends the detected data to the microcontroller, the microcontroller sends the detected data to the LED display screen for display and starts the warning light for warning, and the microcontroller sends the detected data and location information to the background server at the same time;

Step 7: The background server classifies and stores the received data, and performs fuzzy cluster analysis on the accumulated data to obtain a more accurate feature vector interval, and periodically sends the feature vector interval as an updated feature vector sample set to all the data. described microcontroller. The microcontroller makes judgments based on this feature vector sample set to make judgments more accurate.

In a possible implementation manner of this embodiment, in step 2, the process of performing clutter removal processing on the magnetic field data in the Z-axis direction of the geomagnetic sensor is: reading the magnetic field data in the Z-axis direction of the geomagnetic sensor, The incoming data at one moment is used as a reference, and the Kalman filter algorithm is used to compare the incoming data at the present moment, infer the wrong data in the moment data, remove the wrong data caused by irrelevant factors, and extract useful data information.

In a possible implementation manner of this embodiment, in step 4, the specific process of data fusion is as follows: storing the incoming data from the geomagnetic sensor in the first seven bits of an eight-bit binary number pre-created by the single-chip microcomputer; The data of the microwave sensor is put into the eighth bit, and then the eight-bit binary data is converted into hexadecimal data.

Assuming that the types of identifiable driving objects are cars, medium-sized off-road vehicles, and large trucks, the data are divided into three ranges, and through data processing, the output data of the geomagnetic sensor is constrained within the range of 0000-007F, in which the data of small cars is set at 001E or so, medium off-road vehicles are set at around 003C, and large trucks are set at around 0064, the data will not exceed 007F at most. Therefore, when the geomagnetic sensor outputs data in hexadecimal, the eighth bit of the corresponding binary data is idle, which can be used to represent the data output by the microwave sensor. The present invention uses hexadecimal code to determine the expected output, which is expressed as follows according to the identification type :

When there is no moving object, the binary number 00000000 is sent, and the hexadecimal number is represented as 0x00;

When the moving object is only a pedestrian without a vehicle, the binary representation is 10000000—10010000, and the hexadecimal representation is 0x80—0x90;

When the moving object includes a vehicle, the eighth bit is 1, and the first to seventh bits represent vehicle information, that is, vehicle data + 0x80.

In a possible implementation manner of this embodiment, in step 5, the specific process of comparing the fused data with the hexadecimal feature vector samples is as follows: the fused data and the hexadecimal feature vector samples Loop detection is performed, and the fused data is matched with each feature vector sample, and if the matching is successful, the information corresponding to the feature vector sample is returned.

In a possible implementation manner of this embodiment, in step 6, the microcontroller sends the detection data to the LED display screen for display and activates the warning light for warning. The specific process is: the microcontroller analyzes the incoming data from the radar sensor. According to the incoming data, the speed of the moving object and the distance information of the moving object from the corner are calculated, and according to the corresponding type of the moving object, different data is called and sent to the LED display screen for display and the warning light is activated to warn.

In a possible implementation of the present embodiment, in step 7, the specific process that the background server performs fuzzy cluster analysis on the accumulated data is:

Step 71, the background server takes the latest n sets of fused geomagnetic data and microwave data from the database as a set of initial sample universe, and set each set of data as X ₁ , X ₂ ,...,X _m , and each data has m Each index represents its characteristics, then the characteristic attribute data of the i-th classification object in the initial sample universe is X _i1 ',X _i2 ',...,X _im ';

Step 72, using the following formula to standardize the feature attribute data:

in,

取适当的M值，使r_ij在[0,1]中 分散开；Step 73: Establish a fuzzy similarity relationship matrix, and determine the similarity matrix R=(r _ij ) _n×n by common cluster analysis, where r _ij is the similarity coefficient between X _i and X _j , where

Take an appropriate value of M to make r _ij spread out in [0,1];

Step 74: Perform a transitive closure operation on the similarity matrix R to obtain t(R)=R ⁿ , and on this basis, set the confidence level λ to take different values to obtain different clustering results.

Fig. 2 is a schematic block diagram of a system for detecting and processing visual blind spot information of a turning road according to an exemplary embodiment. As shown in FIG. 2 , an embodiment of the present invention provides a system for detecting and processing visual blind spot information on a turning road, including a geomagnetic sensor, a microwave sensor, a radar sensor, a microcontroller, a communication module, and an LED display set on the turning road site. and warning lights, as well as the background server that communicates with the microcontroller remotely through the communication module, the geomagnetic sensor, microwave sensor, radar sensor, communication module, LED display screen and warning light are respectively connected to the microcontroller, the geomagnetic sensor The radar sensor is arranged at the corner of the turning road, and the radar sensor is arranged at the corner of the turning road and the probe of the radar sensor faces the corner of the turning road In the straight lane ahead, the microcontroller, the communication module, the LED display screen and the warning light are set at the corner of the turning road, and the LED display screen is set facing the straight lane in front of the corner of the turning road.

In a possible implementation manner of this embodiment, the geomagnetic sensor adopts the HMC5883L three-axis geomagnetic sensor, which is used to output the detected moving object data by detecting the metal content of the moving object; the microwave sensor adopts the model RCWL-0516 The microwave sensor is used for whether there is a moving object; the radar sensor adopts the TF03 lidar sensor, which is used to detect the speed of the moving object and the distance information to the turning angle.

In a possible implementation manner of this embodiment, the communication module adopts a SIM868 GSM/GPRS/GPS module.

In a possible implementation manner of this embodiment, the detection and processing system is applied to a one-way lane road and/or a two-way lane road.

As shown in FIG. 3 to FIG. 5 , the hardware location layout required to complete the method for detecting and processing the visual blind spot information of the turning road according to the present invention is as follows. It is assumed that the roads on both sides of the corner of the two-way lane road are Road A and Road B, respectively. There are 3 columns, one of which is set at the corner, 2 radar sensors and 1 solar panel are set on the column, where the radar sensor is placed above the column, the radar sensor probe is facing the corresponding lane, and the solar panel is placed on the top of the column. A column is set on the A side of the corner road and the B side of the corner road. The column is equipped with a microcontroller, SIM868GSM/GPRS/GPS module, a display screen, and a circular warning light. The specific position of each component is, the micro-controller The device is placed in a sealed box together with the SIM868 GSM/GPRS/GPS module. The sealed box is set at the top of the column, the circular warning light is set on the upper part of the column, and the display screen is set below the circular warning light. There are geomagnetic sensors and microwave sensors on road A. The geomagnetic sensor is placed farther from the corner than the microwave sensor, and the microwave sensor is placed closer to the corner. The same is true for road B. In order to facilitate installation and later maintenance, the radar sensor is set on the column, and the solar panel is installed on the column where the radar sensor is set, and the solar energy is used for power supply; , warning lights and microcontrollers, etc.

The data detected and processed by the sensors on road A is displayed by the display screen and warning light of road B. Similarly, the data detected and processed by the sensors on road B is displayed by the display screen and warning light of road A.

Taking the traveling object as a vehicle as an example, the detection process of the visual blind spot information of the turning road using the detection processing system of the present invention mainly includes the following steps:

Step 1. The vehicle is first detected by the geomagnetic sensor, the magnetic field data in the Z-axis direction of the geomagnetic sensor is read, the clutter is removed by the Kalman filter algorithm, and useful data information is extracted; using the incoming data at the previous moment as a reference, and Compare the incoming data at the moment, infer the wrong data in the moment data, and remove the wrong data caused by irrelevant factors as much as possible.

Step 2: The geomagnetic sensor stores the acquired effective magnetic field data after filtering in the microcontroller ROM, and waits for data fusion to facilitate the next data processing.

Step 3. After the moving object is detected by the microwave sensor, the microwave sensor uses a microwave oscillator to emit microwaves, and the receiving head of the microwave sensor receives the microwaves reflected by the obstacles.

Step 4. The microwave sensor outputs the signal to the microcontroller ROM for data fusion.

Step 5. The microcontroller compares the fused data with the hexadecimal feature vector samples. If the matching is successful, it will turn on the radar sensor and other equipment to perform the next measurement and display. Otherwise, the data will be discarded and the system will restart. restore to the original state.

Step 6. The radar sensor sends the relevant information to the microcontroller, the microcontroller performs data processing, sends the processed relevant data to the warning light and the LED display, and simultaneously turns on the GPS function of the SIM868GSM/GPRS/GPS module to obtain GPS information. Then use the GPRS function of the SIM868 GSM/GPRS/GPS module to uniformly package the data and send it to the background server.

Step 7. The server classifies and stores the received data, performs fuzzy clustering analysis after accumulating certain data, and obtains a more accurate feature vector interval, and periodically sends the feature vector interval to the microcontroller as an updated feature vector sample Set, the microcontroller makes judgments based on this vector sample set to make judgments more accurate.

Taking the traveling object as an example, the detection process of the visual blind spot information of the turning road using the detection processing system of the present invention mainly includes the following steps:

Step 1. Pedestrians are detected by the geomagnetic sensor first, read the magnetic field data in the Z-axis direction of the geomagnetic sensor, and extract effective data information. The human body has little or no interference with the geomagnetic signal; use the incoming data at the previous moment as a reference. , compare with the incoming data at the present moment, infer the wrong data in the moment data, and remove the wrong data caused by irrelevant factors as much as possible.

Step 2', the microcontroller receives the geomagnetic sensor signal, and will not obtain a valid signal, then go to step 3'.

Step 3', the pedestrian passes through the microwave sensor, the microwave sensor uses a microwave oscillator to send out microwaves, and the receiving head of the microwave sensor receives the microwaves reflected by the obstacle.

Step 4', the microwave sensor output signal is in the microcontroller ROM, carries out data fusion.

Step 5. The microcontroller compares the fused data with the hexadecimal feature vector samples. If the matching is successful, it will turn on the radar sensor and other equipment to perform the next measurement and display. Otherwise, the data will be discarded and the system will restart. restore to the original state.

Step 6. The radar sensor sends the relevant information to the microcontroller, the microcontroller performs data processing, sends the processed relevant data to the warning light and the LED display, and simultaneously turns on the GPS function of the SIM868GSM/GPRS/GPS module to obtain GPS information. Then use the GPRS function of the SIM868 GS/GPRS/GPS module to uniformly package the data and send it to the background server.

Step 7. The server classifies and stores the received data, performs fuzzy clustering analysis after accumulating certain data, and obtains a more accurate feature vector interval, and periodically sends the feature vector interval to the microcontroller as an updated feature vector sample Set, the microcontroller makes judgments based on this sample set to make judgments more accurate.

The specific method of data fusion is: Assuming that the types of identifiable driving objects are cars, medium-sized off-road vehicles, and large trucks, the data is divided into three ranges, and through data processing, the output data of the geomagnetic sensor is constrained within the range of 0x00-007F, where The data of the small car is set at about 0x1E, the medium off-road vehicle is set at about 0x3C, and the large truck is set at about 0x64, then the data will not exceed 0x7F.

The following data is the data transmitted by the geomagnetic sensor:

car:

0x26,0x20,0x23,0x1C,0x1A,0x18,0x1E,0x21,0x29,0x1D,0x1D, 0x25,0x22,0x21,0x19,0x1A,0x1B,0x1F,0x28,0x23,0x24,0x16,0x19,0x21,0x2 0x1D, 0x18, 0x1E, 0x1A, 0x18, 0x24, 0x18, 0x1F.

Medium SUV:

0x42,0x3B,0x3D,0x41,0x45,0x38,0x43,0x40,0x35,0x3B,0x33,0x3F,0x39,0x3C,0x42,0x33,0x45,0x36,0x34,0x40,0x3E,0x38,x35,0x44,0x3D, 0x36, 0x34, 0x3A, 0x3D, 0x40, 0x45, 0x3B, 0x3D.

Large truck:

0x6E, 0x69, 0x6D, 0x62, 0x61, 0x66, 0x5D, 0x64, 0x5F, 0x5E, 0x65, 0x5C, 0x5E, 0x6A, 0x64, 0x5D, 0x6F, 0x5F, 0x6B, 0x60, 0x5C, 0x67, 0x60, 0x5B, 0x68 0x6A, 0x69, 0x63, 0x5C, 0x5D, 0x66, 0x71, 0x61.

After analyzing the data from the geomagnetic sensor, it is found that the data of different models of vehicles are concentrated in different ranges, and the maximum data does not exceed 0x7F. Therefore, when the geomagnetic sensor outputs data in hexadecimal, the corresponding binary data is eighth. The eighth bit can be used to represent the data output by the microwave sensor. According to different driving objects, the corresponding binary data corresponding to the hexadecimal expected output data after fusion is shown in Table 1.

Table 1:

The specific process of comparing the fused data with the sample set is that the fused data and the hexadecimal feature vector samples are cyclically detected to detect which feature vector sample the fused data matches the most, and if the match is successful, it will return the If the information corresponding to the eigenvector samples is not satisfied, it is considered that the data is generated by some irrelevant factors, and the system does not respond to it.

The microcontroller parses the incoming data from the radar sensor, calculates the speed of the moving object and the distance from the corner according to the incoming data, and transfers different data to the display screen and circular warning according to the corresponding type of the object The lights are used to display different shapes and real-time changing speed and distance information.

The following is a description of the fuzzy clustering analysis process by taking three models of small cars, medium-sized off-road vehicles and large trucks as examples. In the initial stage, the fusion data of n groups of geomagnetic sensors and microwave sensors are obtained from the database as a set of initial sample universe, and similar features are extracted from each data as feature vectors.

The sample set is generated according to the feature vector, and the sample set is used as the matching basis for fuzzy clustering analysis. The principle of fuzzy clustering analysis is as follows:

Step 71, the background server takes the latest n sets of fused geomagnetic data and microwave data from the database as a set of initial sample universe, and set each set of data as X ₁ , X ₂ ,...,X _m , and each data has m Each index represents its characteristics, then the characteristic attribute data of the i-th classification object in the initial sample universe is X _i1 ',X _i2 ',...,X _im ';

Step 72, using the following formula to standardize the feature attribute data:

in,

取适当的M值，使r_ij在[0,1]中 分散开；Step 73: Establish a fuzzy similarity relationship matrix, and determine the similarity matrix R=(r _ij ) _n×n by common cluster analysis, where r _ij is the similarity coefficient between X _i and X _j , where

Take an appropriate value of M to make r _ij spread out in [0,1];

Step 74: Perform a transitive closure operation on the similarity matrix R, t(R) is the transitive closure of the fuzzy similarity matrix R, which is a minimum fuzzy equivalent matrix containing R that can be obtained by the square method, and perform n times on R. Square, get t(R)= ^Rn , and on this basis, set the confidence level λ to take different values to obtain different clustering results.

The server periodically sends the results of fuzzy clustering analysis of n groups of data to the microcontroller as the updated eigenvector interval; at the same time, the server stores the results, and continues to store multiple sets of fuzzy clustering results, and stores these results again. Perform fuzzy clustering analysis, and so on, continue to analyze. The purpose is to count the categories in the data feature set in each analysis result, remove the similar and useless categories, and obtain a more accurate feature vector interval after repeated analysis.

The present invention uses two sensors to output data, realizes the electronization of output information, and is more accurate than the traditional wide-angle lens in which the opposite road information is visually detected by human eyes; The data is effectively filtered out; the fusion algorithm is adopted to fuse the data from multiple sensors, which is convenient for data transmission, and only one data transmission can send all the data, which improves the transmission efficiency and reduces the system burden; Different from the traditional threshold detection method, the fuzzy clustering analysis method is used to more accurately and efficiently identify the type information of driving objects, such as pedestrians, cyclists, and different types of vehicles, and accurately mine useful data. This detection method can be used in various places such as urban detours, mountain detours, residential quarters, underground parking lots, etc., and has a wide range of applications.

The above are only the preferred embodiments of the present invention. For those skilled in the art, without departing from the principles of the present invention, several improvements and modifications can be made, and these improvements and modifications are also regarded as the present invention. the scope of protection of the invention.

Claims (6)

1. a detection and processing method for turning road visual blind spot information, is characterized in that, comprises the following steps: Step 1, use the geomagnetic sensor to detect the moving object on the turning road, if a vehicle is detected, send the magnetic field data of the Z-axis direction of the geomagnetic sensor to the microcontroller, otherwise go to step 3; Step 2, the microcontroller performs clutter removal processing on the magnetic field data in the Z-axis direction of the geomagnetic sensor, and stores the effective data information of the geomagnetic sensor; Step 3, use the microwave sensor to detect obstacles, and send the received microwave signal reflected by the obstacle to the microcontroller; Step 4, the microcontroller stores the received microwave sensor data information, and performs data fusion on the received geomagnetic sensor data and microwave sensor data; Step 5, the microcontroller compares the fused data with the hexadecimal feature vector samples, and if the matching is successful, the measurement and display of step 6 are performed, otherwise, return to step 1; Step 6, the radar sensor sends the detected data to the microcontroller, the microcontroller sends the detected data to the LED display screen for display and activates the warning light for warning, and the microcontroller sends the detected data and location information to the background server at the same time; Step 7: The background server classifies and stores the received data, and performs fuzzy cluster analysis on the accumulated data to obtain a more accurate feature vector interval, and periodically sends the feature vector interval as an updated feature vector sample set to all the data. described microcontroller. 2. the detection and processing method of a kind of turning road visual blind spot information according to claim 1 is characterized in that, in step 2, the magnetic field data of the Z-axis direction of the geomagnetic sensor is carried out to the process of clutter removal processing as follows: read Take the magnetic field data in the Z-axis direction of the geomagnetic sensor, take the data passed in at the previous moment as a reference, and use the Kalman filter algorithm to compare with the data passed in at the current moment. The resulting error data is removed, and useful data information is extracted. 3. the detection and processing method of a turning road visual blind spot information according to claim 1, is characterized in that, in step 4, the concrete process of carrying out data fusion is: the data that the geomagnetic sensor is passed in is stored in the single-chip microcomputer to create in advance The first seven bits of an eight-bit binary number, put the data of the microwave sensor into the eighth bit, and then convert the eight-bit binary data into hexadecimal data. 4. the detection processing method of a kind of turning road visual blind spot information according to claim 1, is characterized in that, in step 5, the concrete process that the data after fusion is compared with hexadecimal feature vector sample is: : The fused data and the hexadecimal feature vector samples are cyclically detected, and the fused data is matched with each feature vector sample. If the matching is successful, the information corresponding to the feature vector sample is returned. 5. The method for detecting and processing the visual blind spot information of a turning road according to claim 1, wherein in step 6, the microcontroller sends the detection data to the LED display screen for display and activates the warning light to warn. The specific process is as follows: the microcontroller parses the incoming data from the radar sensor, calculates the speed of the moving object and the distance from the moving object to the corner according to the incoming data, and calls different data transmissions according to the corresponding type of the moving object. Display the LED display and activate the warning light to warn. 6. the detection processing method of a kind of turning road visual blind spot information according to any one of claim 1-5, it is characterized in that, in step 7, the concrete process that background server carries out fuzzy clustering analysis to accumulated data is: : Step 71, the background server takes the latest n sets of fused geomagnetic data and microwave data from the database as a set of initial sample universe, and set each set of data as X ₁ , X ₂ ,..., X _m , and each data has m Each index represents its characteristics, then the characteristic attribute data of the i-th classification object in the initial sample universe is X _i1 ',X _i2 ',...,X _im '; Step 72, using the following formula to standardize the feature attribute data:

in,

Step 73: Establish a fuzzy similarity relationship matrix, and determine the similarity matrix R=(r _ij ) _n×n by common cluster analysis, where r _ij is the similarity coefficient between X _i and X _j , where

Take an appropriate value of M to make r _ij spread out in [0,1]; Step 74: Perform a transitive closure operation on the similarity matrix R to obtain t(R)=R ⁿ , and on this basis, set the confidence level λ to take different values to obtain different clustering results.

Images