CN113340812A - Multispectral-based road surface water ice and snow recognition device - Google Patents

Abstract

The invention relates to a multispectral-based pavement water ice and snow recognition device, which comprises a shell filled in the ground, and a light-emitting diode, a photodiode and a separator which are arranged in the shell, wherein the light-emitting diode is provided with an infrared light diode, a red light diode and a green light diode, the emitting directions of the infrared light diode, the red light diode and the green light diode are all arranged upwards, the receiving direction of the photodiode is arranged upwards, the separator is arranged between the light-emitting diode and the photodiode and used for light-tight isolation between the light-emitting diode and the photodiode, the top surface of the shell is a colorless transparent surface, and the top surface of the shell is flush with the ground. Compared with the prior art, the device has the advantages that the water, ice and snow states of the middle road surface and the states of foreign matters can be identified by utilizing light with three wavelengths, the accuracy rate of identifying the road surface states is improved, meanwhile, the whole device is simple in structure, easy to arrange, install and maintain, small in power consumption, capable of maintaining long-term operation and convenient to use.

Description

Multispectral-based road surface water ice and snow recognition device

Technical Field

The invention relates to the field of traffic road detection, in particular to a recognition device for recognizing water, ice and snow on a road surface.

Background

Due to the accelerated increase of the number of motor vehicles, the driving environment and conditions of roads in China are difficult to improve, the safety problem of the roads is increasingly prominent, and in fact, continental areas become serious disaster areas frequently occurring in global traffic accidents. Research shows that the key objective factor causing traffic accidents is bad road conditions caused by bad weather, and the bad road conditions are the heart of the highway management department in the world. Taking the ice and snow road conditions as an example, the friction coefficient between the vehicle tire and the ground is reduced rapidly due to the ice and snow, so the probability of traffic accidents such as vehicle collision and scraping is greatly increased. In China, most traffic accidents are directly or indirectly related to the rainy, foggy and frosty weather due to the difference of traffic and climate conditions in each province. Therefore, safe driving can be greatly improved in order to provide timely and effective road surface state information for related functional departments and drivers in severe weather environments.

At present, the detection means for the road condition of roads in China is still very limited, wherein the detection means with wider coverage area adopts video monitoring equipment, and also adopts various road surface state sensors, the road surface state sensors can be divided into two types of contact type and non-contact type from the measurement mode, the non-contact type sensors are often installed at two sides of the roads and have larger detection area, however, when the roads encounter bad weather such as strong snow, fog, rain and the like or cobweb and branches and leaves are attached, the optical paths of the sensors are frequently shielded and even polluted, and the identification accuracy cannot be guaranteed. In addition, the non-contact sensor has low working stability and high use and maintenance cost, and is not suitable for large-scale application. The existing contact sensor generally relies on a multi-frequency capacitance measurement principle to judge icing, but misjudgment of drying and icing states occurs frequently; generally, the thickness of the water film is measured by an active microwave radar, but complex refraction and reflection phenomena can occur in an ice layer, and the phase detection is seriously interfered.

Therefore, the present inventors have made extensive studies on the above problems, and as a result, they have made the present invention.

Disclosure of Invention

The invention aims to provide a multispectral road surface water ice and snow recognition device which is accurate in detection and capable of recognizing the states of road surface water ice and snow and foreign matters.

To achieve the above object, the solution of the present invention is as follows: a multispectral-based pavement water ice and snow recognition device comprises a shell filled in the ground, and a light emitting diode, a photodiode and a separator which are arranged in the shell, wherein the light emitting diode is provided with an infrared diode, a red diode and a green diode, the emitting directions of the infrared diode, the red diode and the green diode are all arranged upwards, the receiving direction of the photodiode is arranged upwards, the separator is arranged between the light emitting diode and the photodiode and used for light-tight isolation between the light emitting diode and the photodiode, the top surface of the shell is a colorless light-transmitting surface, and the top surface of the shell is flush with the ground.

The shell is provided with a lower shell and an upper top plate, the lower shell is a hollow shell with an open top surface, the partition is a light-tight partition plate, the light-tight partition plate is positioned in the lower shell and divides the hollow cavity of the lower shell into a first cavity and a second cavity which are transversely arranged and are not communicated, the upper top plate is stacked on the top surface of the lower shell and is tightly matched with the lower shell, the light-emitting diode is positioned in the first cavity, and the photodiode is positioned in the second cavity.

The infrared light diode, the red light diode, the green light diode and the photodiode are arranged in a row at intervals in the transverse direction, and the light emitting diode and the photodiode are arranged on the inner bottom surface of the shell.

The upper top plate is an organic glass plate.

And a communication module connected with the photodiode is arranged in the second chamber.

After the technical scheme is adopted, when the multispectral road surface water ice and snow recognition device is applied, the light emitting diode emits infrared light, red light and green light, and the photodiode receives reflected light. When the road surface state is a dry state, namely the top surface of the shell is not covered with a substance, the light received by the photodiode is the light reflected back to the device by the upper surface of the shell; when the road surface state is water, ice and other light-transmitting substances, the light received by the photodiode comprises light reflected by the upper surface of the shell back to the device and light reflected by the upper surface of the cover back to the device; when the road surface state is snow, sandy soil and other substances which are almost not light-tight, the light emitted to the upper surface of the shell of the device by the light-emitting diode is almost totally reflected and received by the photodiode, the light received by the photodiode is used for calculating the reflectivity, and the calculated reflectivity is compared with the reference reflectivity of the identification device in different road states to obtain the current road surface state. Compared with the prior art, compared with the traditional identification device, the identification device can identify the road surface state and the foreign matter state by utilizing the light with three wavelengths, can improve the accuracy rate of identifying the road surface state, has simple structure, is easy to arrange, install and maintain, has small power consumption, can maintain long-term operation, and is convenient to use.

Drawings

FIG. 1 is a schematic structural diagram of the present invention.

Detailed Description

In order to further explain the technical solution of the present invention, the present invention is explained in detail by the following specific examples.

The invention relates to a multispectral road surface water ice and snow recognition device, as shown in figure 1, which comprises a shell 1, and a light emitting diode, a photodiode 2 and a partition 3 which are arranged in the shell, wherein the light emitting diode is provided with an infrared light diode 41, a red light diode 42 and a green light diode 43, the emitting directions of the infrared light diode 41, the red light diode 42 and the green light diode 43 are all arranged upwards, preferably, the shell 1 is provided with a lower shell 11 and an upper top plate 12, the lower shell 11 is a hollow shell with an open top surface, the lower shell 11 is a square shell, the partition 3 is an opaque partition, the partition 3 is erected in the lower shell 11, the partition 3 is arranged between the light emitting diode and the photodiode, particularly, the partition 3 is arranged in the lower shell to divide the hollow cavity of the lower shell into a first cavity and a second cavity which are transversely arranged and are not communicated with each other, the first cavity is located on the left side, the second cavity is located on the right side, the upper top plate 12 is stacked on the top surface of the lower shell and is closely adhered and matched with the lower shell 11, the top surface of the partition plate 3 is adhered and fixed with the bottom surface of the upper top plate, the bottom surface of the partition plate 3 is adhered and fixed with the inner bottom surface of the lower shell, the upper top plate 12 is preferably an organic glass plate, the infrared diode 41, the red diode 42 and the green diode 43 are transversely arranged in the left first cavity at intervals, the photodiode 2 is located in the second cavity, the receiving direction of the photodiode 2 is arranged upwards, and the infrared diode 41, the red diode 42, the green diode 43 and the photodiode 2 are preferably arranged on the inner bottom surface of the lower shell 11.

When the identification device is applied, a controller and a display screen are further installed in an outer shell of the identification device, the shell 1 is buried on the ground, the top surface of the shell 1 is the same as the ground, an infrared light diode 41, a red light diode 42 and a green light diode 43 are respectively started to respectively and correspondingly emit infrared light, red light and green light, a photodiode 2 respectively receives reflected light, the intensity of the received reflected light is sent to the controller, and the controller processes the reflected light to judge the corresponding road surface state and displays the road surface state by the display screen; when the road surface state is a dry state, i.e. the top surface of the upper top plate 12 is not covered with a substance, the light received by the photodiode 2 is the light reflected back to the device by the upper top plate 12; when the road surface state is water, ice or other light-transmitting substances, the upper top plate 12 is provided with the cover 100, and the light received by the photodiode 2 is reflected back to the device by the upper top plate 12 and reflected back to the device by the upper surface of the cover 100; when the road surface state is snow, sandy soil and other substances which are almost not light-tight, the light emitted by the light emitting diode to the upper surface of the upper top plate 12 is almost totally reflected and received by the photodiode 2, the light received by the photodiode 2 is used for calculating the reflectivity, and the calculated reflectivity is compared with the reference reflectivity set in the controller in different road states to obtain the current road surface state. Compared with the prior art, compared with the traditional identification device, the identification device can improve the accuracy of identifying the road surface state by utilizing the light with three wavelengths, meanwhile, the whole device has simple structure, is easy to arrange, install and maintain, has small power consumption, can maintain long-term operation, and is convenient to use

The multispectral-based road surface water ice and snow detection method adopting the identification device is realized by the following steps:

embedding the multispectral-based pavement water, ice and snow recognition device on the ground, wherein the top surface of the upper top plate 12 is flush with the ground;

step two, calculating the intensity of the reflected light, placing white paper on the top surface of the shell in the step one, starting the light emitting diode and the photodiode, wherein the wavelength of the infrared photodiode is lambda₁Wavelength of red light diode is lambda₂Wavelength of green diode is lambda₃Respectively obtained from photodiodes at a wavelength of λ₁Intensity P of reflected infrared light_{Infrared ray}Wavelength of λ₂Red reflected light intensity P of_{Red wine}Wavelength of λ₃Intensity P of green light reflection_GreenA plurality of operating currents I with linear proportionality are input to the light-emitting diode under specific wavelength_i(i-1, 2, …, n), the photodiode obtains a plurality of reflected light intensities P corresponding to the specific wavelength_i(i ═ 1, 2, …, n), then, at a wavelength λ₁Can obtain multiple infrared reflection intensities

At a wavelength of λ₂Multiple red light reflection intensities

At a wavelength of λ₃Multiple green light reflection intensities

Step three, calculating the intensity coefficient of the reflected light, and utilizing a mathematical statistical method to carry out comparison on the plurality of working currents I in the step two_iAnd the intensity of the reflected light P_iLinear regression to obtain P_i＝β·I_i+ε_i(i-1, 2, …, n), wherein the slope β of the linear regression line is expressed as the reflected light intensity coefficient per unit current, ε_iThe regression coefficient beta is calculated by the least square method for the noise generated by the environment temperature under the unit current, and the least square estimation value of the beta is calculated when the regression coefficient beta is not influenced by the noise such as the environment temperature

Calculating the calculated intensity coefficient of the reflected light

The formula of (1) is as follows:

wherein the content of the first and second substances,

when calculating will

Substituting into the above formula to obtain white paper with wavelength of λ₁Reflection intensity coefficient beta of lower infrared light_N1Will be

Substituting into the above formula to obtain white paper with wavelength of λ₂Reflection intensity coefficient beta of lower red light_N2Will be

Substituting into the above formula to obtain white paper with wavelength of λ₃Reflection intensity coefficient beta of lower green light_N3；

Step four, calculating the light intensity coefficient according to the wavelength lambda of the white paper₁Has a reflectance of r under infrared light_N1At a wavelength of λ₂Red light of reflectance of r_N2At a wavelength of λ₃Has a reflectance of r in green_N3R is_N1、r_N2、r_N3Is a known value according to the coefficient of intensity of the emitted light beta_MEqual to the incident light intensity coefficient beta_NReflectance r of white paper at wavelength_NThe relation of the ratio of the light source to the white paper and the multispectral-based road surface water ice and snow identification device obtained in the third step have the wavelength lambda₁Coefficient of reflected light intensity beta_N1For wavelength λ₂Reflection intensity coefficient beta of lower red light_N2And for λ₃Reflection intensity coefficient beta of lower red light_N3Calculating the wavelength as lambda₁Coefficient of intensity of emergent light

Wavelength of λ₂Coefficient of intensity of emergent light

Wavelength of λ₃Coefficient of intensity of emergent light

Step five, calculating the reference reflectivity under any road surface state, wherein the reflectivity is equal to the ratio of the reflected light intensity coefficient to the emergent light intensity coefficient, and the emergent light intensity coefficient beta_MEqual to the incident light intensity coefficient beta_NReflectance r of white paper at wavelength_NThe relationship of the ratio of the reflection coefficient to the reflection coefficient of the road surface at the current state and the current wavelength

Wherein beta is_iFor the reflected light intensity coefficient detected under the current road surface state, lambda is entered under the current road surface state₁Reflectivity of

Wherein beta is_i1For the current road surface state at a wavelength lambda₁The calculated reflection light intensity coefficient is the wavelength lambda under the current road surface state₂Reflectivity of

Wherein beta is_i2For the weather conditions at a wavelength λ₂The corresponding wavelength lambda of the reflected light intensity coefficient obtained by calculation under the current road surface state₃Reflectivity of

Wherein beta is_i3At a wavelength lambda in the current road surface state₃Calculating the reflected light intensity coefficient according to six road surface states of dryness, water accumulation, icing, snow accumulation, sewage and sandy soil, namely at the wavelength lambda₁The reflectivity r of the dry road surface is obtained_{Drying 1}Reflectivity r of water accumulation road surface_{Accumulated water 1}And the reflectivity r of the frozen road surface_{Freezing 1}Reflectivity r of snow covered road surface_{Accumulated snow 1}Reflectivity r of sewage road surface_{Waste water 1}Reflectivity r of sandy soil pavement_{Sand soil 1}；

For the same reason, at wavelength λ₂The reflectivity r of the dry road surface is obtained_{Drying 2}Reflectivity r of water accumulation road surface_{Accumulated water 2}And the reflectivity r of the frozen road surface_{Freezing 2}Reflectivity r of snow covered road surface_{Accumulated snow 2}Reflectivity r of sewage road surface_{Waste water 2}Reflectivity r of sandy soil pavement_{Sand 2}；

At a wavelength λ₃The reflectivity r of the dry road surface is obtained_{Drying 3}Reflectivity r of water accumulation road surface_{Accumulated water 3}And the reflectivity r of the frozen road surface_{Freezing 3}Reflectivity r of snow covered road surface_{Snow cover 3}Reflectivity r of sewage road surface_{Waste water 3}Reflectivity r of sandy soil pavement_{Sand 3}；

Establishing coordinates of different road surface states to form a training set sample; when the coordinate of each road surface state is established, the reference reflectivity of three wavelengths under the current road surface state calculated in the fifth step and the ratio of every two reflectivities in the three reference reflectivities are respectively used as a characteristic value, and the six characteristic values are recorded as the coordinate under the road surface state, namely the reference coordinate corresponding to a dry road surface, the reference coordinate corresponding to a ponding road surface, the reference coordinate corresponding to an icy road surface, the reference coordinate corresponding to a snow road surface, the reference coordinate corresponding to a sewage road surface and the reference coordinate corresponding to a sandy soil road surface are respectively established;

specifically, r is adopted for establishing the reference coordinate corresponding to the dry road surface_{Drying 1}、r_{Drying 2}、r_{Drying 3}、r_{Drying 1}/r_{Drying 2}、r_{Drying 2}/r_{Drying 3}、、r_{Drying 3}/r_{Drying 1}These six eigenvalues;

r is adopted for establishing reference coordinates corresponding to the ponding road surface_{Accumulated water 1}、r_{Accumulated water 2}、r_{Accumulated water 3}、r_{Accumulated water 1}/r_{Accumulated water 2}、r_{Accumulated water 2}/r_{Accumulated water 3}、r_{Accumulated water 3}/r_{Accumulated water 1}These six eigenvalues;

r is adopted for establishing reference coordinates corresponding to the ponding road surface_{Freezing 1}、r_{Freezing 2}、r_{Freezing 3}、r_{Freezing 1}/r_{Freezing 2}、r_{Freezing 2}/r_{Freezing 3}、r_{Freezing 3}/r_{Freezing 1}These six eigenvalues;

r is adopted for establishing reference coordinates corresponding to the accumulated snow road surface_{Accumulated snow 1}、r_{Accumulated snow 2}、r_{Snow cover 3}、r_{Accumulated snow 1}/r_{Accumulated snow 2}、r_{Accumulated snow 2}/r_{Snow cover 3}、r_{Snow cover 3}/r_{Accumulated snow 1}These six eigenvalues;

r is adopted for establishing the reference coordinate corresponding to the sewage pavement_{Waste water 1}、r_{Waste water 2}、r_{Waste water 3}、r_{Waste water 1}/r_{Waste water 2}、r_{Waste water 2}/r_{Waste water 3}、r_{Waste water 3}/r_{Waste water 1}These six eigenvalues;

r is adopted for establishing the reference coordinate corresponding to the sandy soil pavement_{Sand soil 1}、r_{Sand 2}、r_{Sand 3}、r_{Sand soil 1}/r_{Sand 2}、r_{Sand 2}/r_{Sand 3}、r_{Sand 3}/r_{Sand soil 1}These six eigenvalues;

step seven, classifying the tested road surface states by adopting an knn algorithm, and firstly calculating the current reflection intensity coefficient beta of the tested road surface states during classification_MeasuringAnd calculating the reflectivity r of the measured road surface state by using the emergent light intensity coefficient calculated in the second step and the ratio relation that the reflectivity is equal to the reflected light intensity coefficient and the emergent light intensity coefficient_MeasuringI.e. separately calculating the measured road surface state at the wavelength lambda₁Lower reflectivity r_{Side 1}At a wavelength λ₂Lower reflectivity r_{Side 2}And wavelength lambda₃Lower reflectivity r_{Side 3}Calculating three groups of reflectivity, taking the data of the three groups of reflectivity values and the ratio data between every two reflectivities in the three groups of reflectivity values as six-dimensional coordinates of a measured sample, importing the six-dimensional coordinates into the training set sample in the step four, calculating the distance between the measured sample and each coordinate in the training set sample by adopting a calculation formula of Euclidean distance, and obtaining the road condition weight with the minimum distance value through arrangement of each distance value so as to obtain the current road condition;

in particular, the current reflection intensity coefficient beta at the measured road surface state_MeasuringThe calculation method of (1) and beta in step three_iIn the same way, without being repeated here, r is used_MeasuringIs equal to beta_MeasuringWith the fixed beta calculated in step two_MThe ratio of the reflection coefficient to the reflection coefficient r is the reflection coefficient of the road surface under the covering 100_{Side 1}Reflectivity r_{Side 2}And a reflectivity r_{Side 3}，

In the weight ratio, if r₁≈r_{Drying 1}，r₂≈r_Drying2，r₃≈r_{Drying 3}，r₁/r₂≈r_{Drying 1}/r_{Drying 2}，r₂/r₃≈r_{Drying 2}/r_{Drying 3}，r₃/r₁≈r_{Drying 3}/r_{Drying 1}Judging that the road surface state is a dry state;

if r₁≈r_{Accumulated water 1}，r₂≈r_{Accumulated water 2}，r₃≈r_{Accumulated water 3}，r₁/r₂≈r_{Accumulated water 1}/r_{Accumulated water 2}，r₂/r₃≈r_{Accumulated water 2}/r_{Accumulated water 3}，r₃/r₁≈r_{Accumulated water 3}/r_{Accumulated water 1}Judging that the road surface state is a water accumulation state;

if r₁≈r_{Freezing 1}，r₂≈r_{Freezing 2}，r₃≈r_{Freezing 3}，r₁/r₂≈r_{Freezing 1}/r_{Freezing 2}，r₂/r₃≈r_{Freezing 2}/r_{Freezing 3}，r₃/r₁≈r_{Freezing 3}/r_{Freezing 1}Judging that the road surface state is an icing state;

if r₁≈r_{Accumulated snow 1}，r₂≈r_{Accumulated snow 2}，r₃≈r_{Snow cover 3}，r₁/r₂≈r_{Accumulated snow 1}/r_{Accumulated snow 2}，r₂/r₃≈r_{Accumulated snow 2}/r_{Snow cover 3}，r₃/r₁≈r_{Snow cover 3}/r_{Accumulated snow 1}Judging that the road surface state is a water accumulation state;

if r₁≈r_{Waste water 1}，r₂≈r_{Waste water 2}，r₃≈r_{Waste water 3}，r₁/r₂≈r_{Waste water 1}/r_{Waste water 2}，r₂/r₃≈r_{Waste water 2}/r_{Waste water 3}，r₃/r₁≈r_{Waste water 3}/r_{Waste water 1}Judging that the road surface state is a sewage state;

if r₁≈r_{Sand soil 1}，r₂≈r_{Sand 2}，r₃≈r_{Sand 3}，r₁/r₂≈r_{Sand soil 1}/r_{Sand 2}，r₂/r₃≈r_{Sand 2}/r_{Sand 3}，r₃/r₁≈r_{Sand 3}/r_{Sand soil 1}And judging that the road surface state is a sandy soil state.

In the invention, a processor for receiving and processing data of the photodiode can be arranged in the second chamber, the calculation mode of the processor adopts the calculation mode, the signal output end of the processor is provided with a communication module, and the communication module is used for transmitting the data to a monitoring center to realize remote monitoring of the data.

Or a communication module is arranged in the second chamber, and the output end of the photodiode is electrically connected with the communication module, so that the data of the photodiode is transmitted to a monitoring center in a wireless communication mode, and the monitoring center performs calculation processing to obtain the road surface state at that time.

The above examples and drawings are not intended to limit the scope of the present invention, and any suitable changes or modifications thereof by one of ordinary skill in the art should be considered as not departing from the scope of the present invention.

Claims (5)

1. A multispectral-based pavement water ice and snow recognition device is characterized by comprising a shell filled in the ground, and a light emitting diode, a photodiode and a separator which are arranged in the shell, wherein the light emitting diode is provided with an infrared light diode, a red light diode and a green light diode, the emitting directions of the infrared light diode, the red light diode and the green light diode are all arranged upwards, the receiving direction of the photodiode is arranged upwards, the separator is arranged between the light emitting diode and the photodiode and used for light-tight isolation between the light emitting diode and the photodiode, the top surface of the shell is a colorless light-transmitting surface, and the top surface of the shell is flush with the ground.

2. The multispectral-based road surface water ice and snow identification device according to claim 1, wherein: the shell is provided with a lower shell and an upper top plate, the lower shell is a hollow shell with an open top surface, the partition is a light-tight partition plate, the light-tight partition plate is positioned in the lower shell and divides the hollow cavity of the lower shell into a first cavity and a second cavity which are transversely arranged and are not communicated, the upper top plate is stacked on the top surface of the lower shell and is tightly matched with the lower shell, the light-emitting diode is positioned in the first cavity, and the photodiode is positioned in the second cavity.

3. The multispectral-based road surface water ice and snow identification device according to claim 2, wherein: the infrared light diode, the red light diode, the green light diode and the photodiode are arranged in a row at intervals in the transverse direction, and the light emitting diode and the photodiode are arranged on the inner bottom surface of the shell.

4. The multispectral-based road surface water ice and snow identification device according to claim 1, wherein: the upper top plate is an organic glass plate.

5. The multispectral-based road surface water ice and snow identification device according to claim 2, wherein: and a communication module connected with the photodiode is arranged in the second chamber.

Images