Multi-sensor fusion light rainfall sensitivity improving method and improving device

Abstract

The invention provides a multi-sensor fusion light rainfall sensitivity improving method and improving device, and the method comprises the following steps: responding to a vehicle mode to obtain sensor data, the sensor data comprising foresight sensing data and light rainfall sensing data; performing data processing based on the light rainfall sensing data to obtain a discrimination type; vehicle control is carried out on the basis of the judgment type in combination with a preset executing mechanism, the executing mechanism comprises a vehicle body headlamp and / or a windscreen wiper, and when the judgment type is a conventional type, the executing mechanism is directly responded to carry out vehicle control; and when the discriminating type is a feature type, controlling the execution mechanism based on the foresight sensing data. According to the invention, the collected light and rainfall sensor data and front sensor data are fused, the algorithm precision is improved, the environmental adaptability and accuracy of light and rainfall detection are further improved, the false identification phenomena such as false wiping of a windscreen wiper and false turning-on of a vehicle body headlamp are greatly reduced, and the hardware cost is saved during application.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle intelligent control and data processing, and in particular relates to a method and device for improving the light rainfall sensitivity of multi-sensor fusion. Background Art

[0002] With the development of automobile technology, vehicles will be equipped with various sensors to perceive information inside the vehicle. Currently, the most installed are forward-looking sensors for sensing information of vehicles ahead and optical rain sensors for sensing light and rain outside the vehicle. Among them, the optical rain sensor information is mainly used to control the switch of the vehicle headlights and wipers. The optical rain sensor perceives the external environment information based on the refraction information of light outside the vehicle. In order to obtain relatively accurate optical rain information, the number of sensing components on the sensor will be increased.

[0003] Currently, the existing solutions have the following disadvantages:

[0004] 1) The optical rainfall sensor needs to be equipped with a separate controller, which increases the hardware cost;

[0005] 2) The optical rain sensor senses external light and rainfall information and is easily affected by the vehicle driving environment, such as tunnels, shaded roads, basements, etc., and may suffer from misidentification. Summary of the invention

[0006] In view of the above-mentioned shortcomings of the prior art, an object of the present invention is to provide a method and device for improving the light rainfall sensitivity of multi-sensor fusion, so as to solve the above-mentioned problems.

[0007] In a first aspect, the present invention provides a method for improving the light rainfall sensitivity of multi-sensor fusion, the method comprising the following steps:

[0008] acquiring sensor data in response to the vehicle mode, wherein the sensor data includes forward vision sensor data and light and rain sensor data;

[0009] Based on the light and rainfall sensor data, data processing is performed to obtain a discrimination type, wherein the discrimination types include dark day, non-dark day, quasi-dark day, rainy day, non-rainy day and quasi-rainy day, wherein dark day, non-dark day, rainy day and non-rainy day are conventional types, and quasi-dark day and quasi-rainy day are characteristic types;

[0010] The vehicle is controlled based on the identified type in combination with a preset actuator, the actuator including a vehicle headlight and a windshield wiper, wherein:

[0011] When the determination type is the conventional type, directly responding to the actuator to control the vehicle;

[0012] When the determination type is the characteristic type, the actuator is controlled based on the forward-looking sensor data.

[0013] In a possible implementation of the present application, controlling the actuator based on the forward-looking sensor data includes: inputting the forward-looking sensor data into a preset image algorithm for analysis and processing to obtain an image category, and controlling the vehicle based on the image category in response to the actuator.

[0014] In a possible implementation of the present application, the image categories include non-rainy daytime, rainy daytime, non-rainy dark day, rainy dark day, tunnel / garage and unknown categories, and the vehicle control based on the image category in response to the actuator includes:

[0015] When the vehicle headlights are in the automatic gear, the forward-looking sensor data is analyzed based on a preset image algorithm, wherein if the analysis result is that it is not rainy and dark, rainy and dark, and in a tunnel / garage, the vehicle headlights are controlled to turn on, otherwise they are not turned on;

[0016] When the wiper is in the automatic gear, the forward-looking sensor data is analyzed based on a preset image algorithm, wherein if the analysis result is a rainy day or a rainy night, the wiper is controlled to be turned on, otherwise not turned on.

[0017] In a possible implementation of the present application, the response vehicle mode is used to obtain sensor data, wherein the sensor data includes forward vision sensor data and light rainfall sensor data, specifically including:

[0018] When the vehicle is powered on and started, the forward vision sensor device and the light and rain sensor device are turned on;

[0019] Acquiring the forward-looking sensing data based on the forward-looking sensing device, wherein the forward-looking sensing device includes a camera;

[0020] Light and rain sensing data are acquired based on a light and rain sensing device, wherein the light and rain sensing device comprises a light and rain sensor, and the light and rain sensing data comprises light data and rainfall data.

[0021] In a possible implementation of the present application, data processing is performed based on light data to obtain dark sky, non-dark sky and quasi-dark sky, specifically including:

[0022] When the number of continuously collected light data values ​​is greater than or equal to the first data value number threshold, and the interval time of collecting light is greater than or equal to the first interval threshold, the light data is processed based on the light perception algorithm, wherein:

[0023] When the light value is less than or equal to the dark sky light value threshold, and the dark sky light threshold number is greater than or equal to the first percentile of the number of continuously collected light data values, the processing result output based on the light perception algorithm is dark sky;

[0024] When the light value is greater than the dark sky light value threshold and less than or equal to the quasi-dark sky light value threshold, and the quasi-dark sky light threshold number is greater than or equal to the first percentile of the number of continuously collected light data values, the processing result output based on the light perception algorithm is quasi-dark sky;

[0025] Otherwise, the processing result output based on the light perception algorithm is that it is not dark.

[0026] In a possible implementation of the present application, data processing is performed based on rainfall data to obtain rainy days, non-rainy days and quasi-rainy days, specifically including:

[0027] When the number of continuously collected rainfall data values ​​is greater than or equal to the second data value number threshold, and the interval time of collecting rainfall is greater than or equal to the second interval threshold, the rainfall data is processed based on the rainfall sensing algorithm, wherein:

[0028] When the rainfall value is less than or equal to the rainy day rainfall value threshold, and the number of rainy day rainfall thresholds is greater than or equal to the second percentile of the number of continuously collected rainfall data values, the processing result output based on the rainfall sensing algorithm is rainy day;

[0029] When the rainfall value is greater than the rainy day rainfall value threshold and less than or equal to the quasi-rainy day rainfall value threshold, and the number of quasi-rainy day rainfall thresholds is greater than or equal to the second percentile of the number of continuously collected rainfall data values, the processing result output based on the rainfall sensing algorithm is a quasi-rainy day;

[0030] Otherwise, the processing result output based on the rainfall sensing algorithm is that it is not a rainy day.

[0031] In a possible implementation of the present application, when the type is determined to be the conventional type, the vehicle control is performed in direct response to the actuator, specifically including:

[0032] When the vehicle headlights are in the automatic gear, the vehicle headlights are automatically turned on when it is dark based on the output processing result of the light sensing algorithm; and the vehicle headlights are not turned on when it is not dark based on the output processing result of the light sensing algorithm;

[0033] When the wiper is in the automatic gear, the wiper is controlled to open automatically when the processing result of the rain sensing algorithm output is rainy; and when the processing result of the rain sensing algorithm output is not rainy, the wiper is controlled not to open.

[0034] In a second aspect, the present invention provides a multi-sensor fusion light rainfall sensitivity improvement device, which is applied to any of the multi-sensor fusion light rainfall sensitivity improvement methods described above, wherein the device comprises:

[0035] A sensor unit, used to obtain sensor data, the sensor unit includes a forward-looking sensor and an optical rainfall sensor, wherein the sensor data includes forward-looking sensing data and optical rainfall sensing data, the forward-looking sensor is used to obtain the forward-looking sensing data, and the optical rainfall sensor is used to obtain the optical rainfall sensing data;

[0036] A controller unit, used for performing data processing based on the light and rainfall sensing data to obtain a discrimination type, the controller unit comprising an image algorithm module, a light and rainfall algorithm module and a control algorithm module, wherein the discrimination types include dark sky, non-dark sky, quasi-dark sky, rainy day, non-rainy day and quasi-rainy day, wherein dark sky, non-dark sky, rainy day and non-rainy day are conventional types, and quasi-dark sky and quasi-rainy day are characteristic types;

[0037] An actuator unit is used to control the vehicle based on the discrimination type in combination with a preset actuator, the actuator unit including a body controller, body headlights and / or wipers, wherein when the discrimination type is the conventional type, the vehicle is controlled in direct response to the actuator; when the discrimination type is the characteristic type, the actuator is controlled based on the forward-looking sensor data.

[0038] In a possible implementation of the present application, the light and rain sensor data include light data and rain data, and the light and rain algorithm module includes a light perception algorithm and a rain perception algorithm, wherein the light perception algorithm processes the light data to obtain dark day, non-dark day and quasi-dark day results, and the rain perception algorithm processes the rain data to obtain rainy day, non-rainy day and quasi-rainy day, and the image algorithm analyzes and processes the forward-looking sensor data to obtain image categories, wherein the image categories include non-rainy daytime, rainy daytime, non-rainy dark day, rainy dark day, tunnel / garage and unknown categories.

[0039] In a possible implementation of the present application, when the vehicle headlights are in the automatic gear, based on the light perception algorithm output result that it is dark, the control algorithm module controls the vehicle headlights to automatically turn on through the vehicle body controller; when the output result is non-dark, the control algorithm module controls the vehicle headlights not to turn on through the vehicle body controller; when the output result is quasi-dark, the forward-looking sensor data is analyzed based on the image algorithm, wherein if the analysis result is a tunnel / garage, the control algorithm module controls the vehicle headlights to turn on through the vehicle body controller, otherwise not to turn on; and when the wipers are in the automatic gear, based on the rain perception output result that it is rainy, the control algorithm module controls the wipers to automatically turn on through the vehicle body controller; when the output processing result is non-rainy, the control algorithm module controls the wipers not to turn on through the vehicle body controller; when the output result is quasi-rainy, the forward-looking sensor data is analyzed based on the image algorithm, wherein if the analysis result is rainy daytime or rainy dark, the control algorithm module controls the wipers to turn on through the vehicle body controller, otherwise not to turn on.

[0040] As described above, the method and device for improving the light and rain sensitivity of multi-sensor fusion described in the present invention have the following beneficial effects: the collected light and rain sensor data and the front sensor data are fused to improve the algorithm accuracy, thereby improving the environmental adaptability and accuracy of light and rain detection, greatly reducing the misidentification phenomena such as wiper miswiping and vehicle headlight misopening, and when the light and rain sensor is used, the MCU of the front sensor controller unit is reused to realize algorithm detection, which can reduce hardware costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic diagram showing the steps of a method for improving the light rainfall sensitivity of multi-sensor fusion according to an embodiment of the present invention;

[0042] Figure 2 It is a schematic diagram showing the image algorithm processing in one embodiment of the method for improving the light rainfall sensitivity by multi-sensor fusion of the present invention;

[0043] Figure 3 It is a schematic diagram showing the flow of a method for improving the light rainfall sensitivity of multi-sensor fusion in one embodiment of the present invention;

[0044] Figure 4 It is a schematic diagram of power-on control of a method for improving the sensitivity of light rainfall by multi-sensor fusion according to an embodiment of the present invention;

[0045] Figure 5 It is a schematic diagram of a light perception algorithm in an embodiment of a method for improving light rainfall sensitivity by multi-sensor fusion of the present invention;

[0046] Figure 6It is a schematic diagram of a rain sensing algorithm in one embodiment of a method for improving light rain sensitivity by multi-sensor fusion of the present invention;

[0047] Figure 7 It is a schematic structural diagram of a multi-sensor fusion light rainfall sensitivity enhancement device in one embodiment of the present invention. DETAILED DESCRIPTION

[0048] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0049] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0050] The present invention integrates the environmental information perceived by the forward-looking sensor with the light and rain perception information to jointly judge the external light information of the vehicle, thereby more accurately controlling the operation of the vehicle headlights and wipers. The light and rain sensor reuses the forward-looking sensor controller unit MCU to implement algorithm detection, which can reduce hardware costs; and reuses the forward-looking sensor information to judge the driving environment, which can improve the sensitivity of the light and rain sensor.

[0051] The technical solutions in the embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0052] Specifically, see Figure 1 In one embodiment of the invention, the method for improving the light rainfall sensitivity of multi-sensor fusion of the present invention comprises the following steps:

[0053] Step S102, acquiring sensor data in response to the vehicle mode, wherein the sensor data includes forward vision sensor data and light and rain sensor data;

[0054] Step S104, performing data processing based on the light and rainfall sensing data to obtain a discrimination type, wherein the discrimination types include dark sky, non-dark sky, quasi-dark sky, rainy day, non-rainy day and quasi-rainy day, wherein dark sky, non-dark sky, rainy day and non-rainy day are conventional types, and quasi-dark sky and quasi-rainy day are characteristic types;

[0055] Step S106, controlling the vehicle based on the identified type and a preset actuator, wherein the actuator includes a vehicle headlight and / or a windshield wiper;

[0056] Step S108, when the determination type is the conventional type, directly respond to the actuator to control the vehicle; when the determination type is the characteristic type, control the actuator based on the forward-looking sensor data.

[0057] It should be noted that, in this embodiment, after the vehicle is powered on, the forward-looking sensor device and the light and rain sensor device will be started synchronously to obtain corresponding sensor data. Correspondingly, the sensor data obtained includes forward-looking sensor data and light and rain sensor data, and then data processing is performed based on the light and rain sensor data to obtain a discrimination type, and the discrimination type includes dark sky, non-dark sky, quasi-dark sky, rainy day, non-rainy day and quasi-rainy day, among which dark sky, non-dark sky, rainy day and non-rainy day are conventional types, and quasi-dark sky and quasi-rainy day are characteristic types. Since the purpose of the description in this embodiment is to integrate multiple sensors The combination is specifically the fusion of the forward-looking sensor data and the light and rain sensor data. Therefore, it is necessary to combine different discrimination types for processing in order to combine with the preset actuators for vehicle control. The actuators include vehicle headlights and / or wipers, wherein the discrimination types include conventional types and characteristic types. Specifically, when the discrimination type is the conventional type, the vehicle control is performed in direct response to the actuators; when the discrimination type is the characteristic type, the actuators are controlled based on the forward-looking sensor data, wherein the specific control method will be described in detail in the subsequent instructions.

[0058] Furthermore, in one embodiment of the invention, controlling the actuator based on the forward-looking sensor data includes: inputting the forward-looking sensor data into a preset image algorithm for analysis and processing to obtain an image category, and controlling the vehicle based on the image category in response to the actuator.

[0059] It should be noted that, in this embodiment, a deep learning algorithm is used to classify and judge the image information collected by the forward-looking sensor, and the external environment of the vehicle is divided into the following categories: non-rainy daytime, rainy daytime, non-rainy nighttime, rainy nighttime, tunnel (including garage), and unknown (Unknow) category. That is, the corresponding image categories include non-rainy daytime, rainy daytime, non-rainy nighttime, rainy nighttime, tunnel / garage and unknown category. Specifically, after the car is powered on, the forward-looking sensor starts to continuously collect images, wherein the operation of the image algorithm depends on the light rainfall algorithm. When the light rainfall algorithm outputs an uncertain result (quasi-rainy day and quasi-dark day), the image algorithm is used to process the images continuously collected by the forward-looking sensor. Specifically, the purpose of using the image algorithm to process data is to use deep learning technology to perform pattern recognition when the light rainfall sensor algorithm obtains an uncertain result, thereby realizing secondary judgment of the result and reducing the error control rate of the light rainfall sensor.

[0060] Specifically, Figure 2 As shown, the forward-looking sensor obtains the video stream I t ∈R H*W*C ,t=0,...T, when the light rain sensor algorithm (light sensing algorithm and rain sensing algorithm) obtains uncertain results, collect I t The image is input into the image algorithm for processing, and the image classification result is output using the deep learning algorithm. The deep learning algorithm module M class ResNet-50 is used, and its output layer is a fully connected layer + Sigmoid activation function to obtain the category feature vector F, and then perform ArgMax operation and confidence Conf class After filtering, the final category C of the image is obtained (non-rainy daytime, rainy daytime, non-rainy nighttime, rainy nighttime, tunnel / garage, unknown category).

[0061] Further, in one embodiment of the invention, the vehicle control based on the conventional type directly responding to the actuator specifically includes:

[0062] When the vehicle headlights are in the automatic gear, the vehicle headlights are controlled to be turned on when it is dark based on the output processing result of the light sensing algorithm; and the vehicle headlights are controlled not to be turned on when it is not dark based on the output processing result of the light sensing algorithm;

[0063] When the wiper is in the automatic gear, the wiper is controlled to be turned on when the processing result of the rain sensing algorithm output is rainy; and the wiper is controlled not to be turned on when the processing result of the rain sensing algorithm output is not rainy.

[0064] Further, in one embodiment of the invention, controlling the vehicle in response to the actuator based on the image category includes:

[0065] When the vehicle headlights are in the automatic gear position, the forward-looking sensor data is analyzed based on a preset image algorithm, wherein if the analysis result indicates a tunnel / garage, the vehicle headlights are controlled to be turned on, otherwise not turned on;

[0066] When the wiper is in the automatic gear, the forward-looking sensor data is analyzed based on a preset image algorithm, wherein if the analysis result is a rainy day or a rainy night, the wiper is controlled to be turned on, otherwise not turned on.

[0067] It should be noted that, in this embodiment, Figure 3 As shown, the content of controlling the body headlights and wipers is specifically explained, wherein, when the vehicle is started, the forward-looking sensor and the light and rain sensor work, when the body headlight switch is in the automatic gear (Auto gear), the light and rain sensor collects light data, and processes the light data using the light perception algorithm. If it is judged to be dark, the body headlights are controlled to be turned on; if it is judged to be not dark, the body headlights are controlled not to be turned on; if it is judged to be quasi-dark, the forward-looking sensor data is analyzed by the image algorithm. If the analysis result is dark in rainy weather, dark in non-rainy weather, and tunnel / garage, the body headlights are controlled to be turned on, otherwise they are not turned on.

[0068] Furthermore, when the vehicle is started, the forward-looking sensor and the optical rain sensor work. When the wiper is in the automatic position (Auto position), the optical rain sensor collects rainfall data and processes the rainfall data using a rainfall sensing algorithm. If it is judged to be a rainy day, the wipers are controlled to turn on; if it is judged to be a non-rainy day, the wipers are controlled not to turn on; if it is judged to be a quasi-rainy day, an image algorithm is used to analyze the forward-looking sensor data. If the analysis result shows that it is raining during the day or raining at night, the wipers are controlled to turn on, otherwise they are not turned on.

[0069] Further, in one embodiment of the invention, the response vehicle mode is used to obtain sensor data, wherein the sensor data includes forward vision sensor data and light and rain sensor data, specifically including:

[0070] When the vehicle is powered on and started, the forward vision sensor device and the light and rain sensor device are turned on;

[0071] Acquiring the forward-looking sensing data based on the forward-looking sensing device, wherein the forward-looking sensing device includes a camera;

[0072] Light and rain sensing data are acquired based on a light and rain sensing device, wherein the light and rain sensing device comprises a light and rain sensor, and the light and rain sensing data comprises light data and rainfall data.

[0073] It should be noted that, in this embodiment, Figure 4As shown, when the vehicle is powered on and started, the forward-looking sensor device and the light and rain amount sensor device are turned on; when the vehicle is not powered on and started, the forward-looking sensor device and the light and rain amount sensor device are also turned off accordingly; further, when the forward-looking sensor device and the light and rain amount sensor device are started, data are automatically collected, wherein the forward-looking sensor data is obtained based on the forward-looking sensor device, wherein the forward-looking sensor device includes a camera, which is specifically installed at the windshield of the vehicle, and senses environmental information for judging driving functions and driving environments; and light and rain amount sensor data is obtained based on the light and rain amount sensor device, wherein the light and rain amount sensor device includes a light and rain amount sensor, and the light and rain amount sensor data includes light data and rainfall data, and accordingly, the light and rain amount sensor can also be installed at the windshield of the vehicle to sense changes in light outside the vehicle, and the sensed information is used to control the vehicle headlights and wipers to work.

[0074] Further, in one embodiment of the invention, data processing is performed based on light data to obtain dark sky, non-dark sky and quasi-dark sky, specifically including:

[0075] When the number of continuously collected light data values ​​is greater than or equal to the first data value number threshold, and the interval time of collecting light is greater than or equal to the first interval threshold, the light data is processed based on the light perception algorithm, wherein:

[0076] When the light value is less than or equal to the dark sky light value threshold, and the dark sky light threshold number is greater than or equal to the first percentile of the number of continuously collected light data values, the processing result output based on the light perception algorithm is dark sky;

[0077] When the light value is greater than the dark sky light value threshold and less than or equal to the quasi-dark sky light value threshold, and the quasi-dark sky light threshold number is greater than or equal to the first percentile of the number of continuously collected light data values, the processing result output based on the light perception algorithm is quasi-dark sky;

[0078] Otherwise, the processing result output based on the light perception algorithm is that it is not dark.

[0079] It should be noted that, in this embodiment, Figure 5 As shown, the number of continuously collected light data values ​​is Sum n , the first data value quantity threshold is S s , the interval time of collecting light is T j , the first interval threshold is T s , the light value is L s , the dark sky light threshold number is Sum s , the threshold number of dark sky light is L l , the first percentile is "80%", and the threshold value of quasi-darkness light is L h , the threshold number of quasi-dark sky light is Sumz Specifically, the light data is used to determine whether it is necessary to turn on the headlights of the vehicle body. The weather is divided into non-dark days, quasi-dark days, and dark days. Quasi-dark days correspond to the period from the beginning of the darkening to the complete darkening, and from the moment of entering the tunnel / garage to the complete entry; dark days correspond to the period from the complete darkening / low visibility, and the vehicle completely entering the tunnel / garage. Non-dark days correspond to the weather other than quasi-dark days and dark days. When the number of continuously collected light data values ​​satisfies the following formula (1), the data in this batch is sent to the light perception algorithm to determine the weather.

[0080]

[0081] Among them, Sum n is the number of continuously collected light data values, S s is the first data value quantity threshold, for example, "10-20" times, T j is the interval time of collecting light, specifically the time since the last light data collection in the previous batch, T s is the first interval threshold, specifically the threshold between the intervals between two batches of light data collection, for example, "0.3-0.6" seconds;

[0082] It is judged as dark sky when the following formula (2) is satisfied, where:

[0083]

[0084] Among them, L l is the light value threshold at night (determined by vehicle calibration), for example, "0.65-0.75", L s is the light value, Sum s is the dark sky light threshold number, Sum n The number of light data values ​​collected continuously;

[0085] When the following formula (3) is satisfied, it is judged as quasi-dark sky, where:

[0086]

[0087] Among them, L l is the light value threshold at night (determined based on vehicle calibration), L h is the quasi-dark sky light value threshold (determined based on vehicle calibration), L s is the light value, Sum z is the threshold number of quasi-dark sky light, Sum n The number of light data values ​​collected continuously.

[0088] Further, in one embodiment of the invention, data processing is performed based on rainfall data to obtain rainy days, non-rainy days and quasi-rainy days, specifically including:

[0089] When the number of continuously collected rainfall data values ​​is greater than or equal to the second data value number threshold, and the interval time of collecting rainfall is greater than or equal to the second interval threshold, the rainfall data is processed based on the rainfall sensing algorithm, wherein:

[0090] When the rainfall value is less than or equal to the rainy day rainfall value threshold, and the number of rainy day rainfall thresholds is greater than or equal to the second percentile of the number of continuously collected rainfall data values, the processing result output based on the rainfall sensing algorithm is rainy day;

[0091] When the rainfall value is greater than the rainy day rainfall value threshold and less than or equal to the quasi-rainy day rainfall value threshold, and the number of quasi-rainy day rainfall thresholds is greater than or equal to the second percentile of the number of continuously collected rainfall data values, the processing result output based on the rainfall sensing algorithm is a quasi-rainy day;

[0092] Otherwise, the processing result output based on the rainfall sensing algorithm is that it is not a rainy day.

[0093] It should be noted that, in this embodiment, Figure 6 As shown, the number of continuously collected rainfall data values ​​is Sum m , the second data value quantity threshold is R s , the interval time for collecting rainfall is T i , the second interval threshold is T r , the rainfall value is R r , the rainy day rainfall threshold is R l , the number of rainy day rainfall thresholds is Sum r , the second percentile is "75%", and the threshold of rainfall value for quasi-rainy days is R h , the number of quasi-rainy day rainfall thresholds is Sum y Specifically, rainfall data is used to determine whether the wipers need to be turned on. The weather is divided into rainy days (moderate rain or above), quasi-rainy days (light rain to moderate rain, non-rainy days with high humidity) and non-rainy days (low humidity, no rain). Furthermore, when the continuously collected rainfall data satisfies the following formula (4), the batch of data is sent to the rainfall sensing algorithm to determine the weather, where:

[0094]

[0095] Among them, Sum m is the number of continuously collected rainfall data values, R s is the second data value quantity threshold, for example, "10-20" times, T iis the interval time of rainfall collection, specifically the time since the last rainfall data collection in the previous batch, T r is the second interval threshold, specifically the threshold between the collection intervals of two batches of rainfall data, for example, "0.3-0.6" seconds;

[0096] When the data satisfies the following formula (5), it is judged as a rainy day, where:

[0097]

[0098] Among them, R l is the rainy day rainfall threshold (determined based on vehicle calibration), R r is the rainfall value, Sum r is the rainfall threshold number for rainy days, Sum m The number of continuously collected rainfall data values;

[0099] It is judged as a quasi-rainy day when the following formula (6) is satisfied, where:

[0100]

[0101] Among them, R l is the rainy day rainfall threshold (determined based on vehicle calibration), R r is the rainfall value, R h is the threshold of rainfall value on rainy days (determined according to vehicle calibration), Sum y is the number of quasi-rainy day rainfall thresholds, Sum m The number of continuously collected rainfall data values.

[0102] Further, in one embodiment of the invention, the vehicle control based on the conventional type directly responding to the actuator specifically includes:

[0103] When the vehicle headlights are in the automatic gear, the vehicle headlights are automatically turned on when it is dark based on the output processing result of the light sensing algorithm; and the vehicle headlights are not turned on when it is not dark based on the output processing result of the light sensing algorithm;

[0104] When the wiper is in the automatic gear, the wiper is controlled to open automatically when the processing result of the rain sensing algorithm output is rainy; and when the processing result of the rain sensing algorithm output is not rainy, the wiper is controlled not to open.

[0105] See also Figure 7 In one embodiment, a multi-sensor fusion light rainfall sensitivity enhancement device is provided in this embodiment, and the device includes:

[0106] A sensor unit, used to obtain sensor data, the sensor unit includes a forward-looking sensor and an optical rainfall sensor, wherein the sensor data includes forward-looking sensing data and optical rainfall sensing data, the forward-looking sensor is used to obtain the forward-looking sensing data, and the optical rainfall sensor is used to obtain the optical rainfall sensing data;

[0107] A controller unit, used for performing data processing based on the light and rainfall sensing data to obtain a discrimination type, the controller unit comprising an image algorithm module, a light and rainfall algorithm module and a control algorithm module, wherein the discrimination types include dark sky, non-dark sky, quasi-dark sky, rainy day, non-rainy day and quasi-rainy day, wherein dark sky, non-dark sky, rainy day and non-rainy day are conventional types, and quasi-dark sky and quasi-rainy day are characteristic types;

[0108] An actuator unit is used to control the vehicle based on the discrimination type in combination with a preset actuator, the actuator unit including a body controller, body headlights and / or wipers, wherein when the discrimination type is the conventional type, the vehicle is controlled in direct response to the actuator; when the discrimination type is the characteristic type, the actuator is controlled based on the forward-looking sensor data.

[0109] It should be noted that, in this embodiment, the sensor unit includes a forward-looking sensor and a light and rain sensor, the forward-looking sensor is used to obtain the forward-looking sensing data, the light and rain sensor is used to obtain the light and rain sensing data, the light and rain sensing data includes light data and rainfall data, the light and rain algorithm module includes a light perception algorithm and a rainfall perception algorithm, wherein the light perception algorithm processes the light data to obtain dark day, non-dark day and quasi-dark day results, the rainfall perception algorithm processes the rainfall data to obtain rainy day, non-rainy day and quasi-rainy day, the image algorithm analyzes and processes the forward-looking sensing data to obtain image categories, wherein the image categories include non-rainy daytime, rainy daytime, non-rainy day dark day, rainy day dark day, tunnel / garage and unknown categories.

[0110] The controller unit includes an image algorithm module for processing forward-looking sensor data collected by the forward-looking sensor, and a light-rainfall algorithm module for processing light data and rainfall data, wherein the light-rainfall algorithm module includes a light perception algorithm and a rainfall perception algorithm, and a control algorithm module for fusing the image algorithm recognition results and the light-rainfall algorithm results, comprehensively judging and outputting signals for switching the body headlights and wipers; and the actuator unit includes a body controller, body headlights and wipers, the body controller receives the control signal of the control algorithm module, and controls the body headlights and wipers according to the corresponding control signal, wherein the body headlights include the front headlights and the rear headlights, and the body headlights turn on or off the body headlights and wipers when receiving the transmission signal of the body controller. The wipers in this embodiment are only the front wipers of the body, which are turned on or off by receiving the transmission signal of the body controller.

[0111] In the several embodiments provided by the present invention, it should be understood that the disclosed system, device or method can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of modules / units is only a logical function division, and there may be other division methods in actual implementation, such as multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules or units, which can be electrical, mechanical or other forms.

[0112] The modules / units described as separate components may or may not be physically separated, and the components displayed as modules / units may or may not be physical modules, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules / units may be selected according to actual needs to achieve the purpose of the embodiments of the present invention. For example, the functional modules / units in the various embodiments of the present invention may be integrated into one processing module, or each module / unit may exist physically separately, or two or more modules / units may be integrated into one module / unit.

[0113] Those of ordinary skill in the art should further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0114] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A method for improving the light rainfall sensitivity of multi-sensor fusion, characterized in that: include: acquiring sensor data in response to the vehicle mode, wherein the sensor data includes forward vision sensor data and light and rain sensor data; Based on the light and rainfall sensor data, data processing is performed to obtain a discrimination type, wherein the discrimination types include dark day, non-dark day, quasi-dark day, rainy day, non-rainy day and quasi-rainy day, wherein dark day, non-dark day, rainy day and non-rainy day are conventional types, and quasi-dark day and quasi-rainy day are characteristic types; The vehicle is controlled based on the identified type in combination with a preset actuator, the actuator including a vehicle headlight and / or a windshield wiper, wherein: When the determination type is the conventional type, directly responding to the actuator to control the vehicle; When the determination type is the characteristic type, the actuator is controlled based on the forward-looking sensor data.

2. The method for improving light rainfall sensitivity by multi-sensor fusion according to claim 1, characterized in that: Controlling the actuator based on the forward-looking sensor data includes: inputting the forward-looking sensor data into a preset image algorithm for analysis and processing to obtain an image category, and controlling the vehicle in response to the actuator based on the image category.

3. The method for improving light rainfall sensitivity by multi-sensor fusion according to claim 2, characterized in that: The image categories include non-rainy daytime, rainy daytime, non-rainy dark daytime, rainy dark daytime, tunnel / garage and unknown categories, and the vehicle control based on the image category in response to the actuator includes: When the vehicle headlights are in the automatic gear, the forward-looking sensor data is analyzed based on a preset image algorithm, wherein if the analysis result is that it is not rainy and dark, rainy and dark, and in a tunnel / garage, the vehicle headlights are controlled to turn on, otherwise they are not turned on; When the wiper is in the automatic gear, the forward-looking sensor data is analyzed based on a preset image algorithm, wherein if the analysis result is a rainy day or a rainy night, the wiper is controlled to be turned on, otherwise not turned on.

4. The method for improving light rainfall sensitivity by multi-sensor fusion according to claim 1, characterized in that: The responding vehicle mode to obtain sensor data, wherein the sensor data includes forward vision sensor data and light and rain sensor data, specifically includes: When the vehicle is powered on and started, the forward vision sensor device and the light and rain sensor device are turned on; Acquiring the forward-looking sensing data based on the forward-looking sensing device, wherein the forward-looking sensing device includes a camera; Light and rain sensing data are acquired based on a light and rain sensing device, wherein the light and rain sensing device comprises a light and rain sensor, and the light and rain sensing data comprises light data and rainfall data.

5. The method for improving light rainfall sensitivity by multi-sensor fusion according to claim 4, characterized in that: Data processing based on light data is used to obtain dark sky, non-dark sky and quasi-dark sky, including: When the number of continuously collected light data values ​​is greater than or equal to a first data value number threshold, and the interval time of collecting light is greater than or equal to the first interval threshold, the light data is processed based on the light perception algorithm, wherein when the light value is less than or equal to the dark sky light value threshold, and the dark sky light threshold number is greater than or equal to the first percentile of the number of continuously collected light data values, the processing result output based on the light perception algorithm is dark sky; When the light value is greater than the dark sky light value threshold and less than or equal to the quasi-dark sky light value threshold, and the quasi-dark sky light threshold number is greater than or equal to the first percentile of the number of continuously collected light data values, the processing result output based on the light perception algorithm is quasi-dark sky; Otherwise, the processing result output based on the light perception algorithm is that it is not dark.

6. The method for improving light rainfall sensitivity by multi-sensor fusion according to claim 4, characterized in that: Data processing based on rainfall data is used to obtain rainy days, non-rainy days and quasi-rainy days, including: When the number of continuously collected rainfall data values ​​is greater than or equal to the second data value number threshold, and the interval time of collecting rainfall is greater than or equal to the second interval threshold, the rainfall data is processed based on the rainfall sensing algorithm, wherein, when the rainfall value is less than or equal to the rainy day rainfall value threshold, and the rainy day rainfall threshold number is greater than or equal to the second percentile of the number of continuously collected rainfall data values, the processing result output based on the rainfall sensing algorithm is rainy day; When the rainfall value is greater than the rainy day rainfall value threshold and less than or equal to the quasi-rainy day rainfall value threshold, and the number of quasi-rainy day rainfall thresholds is greater than or equal to the second percentile of the number of continuously collected rainfall data values, the processing result output based on the rainfall sensing algorithm is a quasi-rainy day; Otherwise, the processing result output based on the rainfall sensing algorithm is that it is not a rainy day.

7. The method for improving light rainfall sensitivity by multi-sensor fusion according to claim 5, characterized in that: When the type is determined to be the conventional type, the vehicle control is performed in direct response to the actuator, specifically including: When the vehicle headlights are in the automatic gear, the vehicle headlights are controlled to be turned on when it is dark based on the output processing result of the light sensing algorithm; and the vehicle headlights are controlled not to be turned on when it is not dark based on the output processing result of the light sensing algorithm; When the wiper is in the automatic gear, the wiper is controlled to be turned on when the processing result of the rain sensing algorithm output is rainy; and the wiper is controlled not to be turned on when the processing result of the rain sensing algorithm output is not rainy.

8. A multi-sensor fusion light rainfall sensitivity enhancement device, characterized in that: The method for improving light rainfall sensitivity of multi-sensor fusion applied to any one of claims 1 to 7, wherein the device comprises: A sensor unit, used to obtain sensor data, the sensor unit includes a forward-looking sensor and an optical rainfall sensor, wherein the sensor data includes forward-looking sensing data and optical rainfall sensing data, the forward-looking sensor is used to obtain the forward-looking sensing data, and the optical rainfall sensor is used to obtain the optical rainfall sensing data; A controller unit, used for performing data processing based on the light and rainfall sensing data to obtain a discrimination type, the controller unit comprising an image algorithm module, a light and rainfall algorithm module and a control algorithm module, wherein the discrimination types include dark sky, non-dark sky, quasi-dark sky, rainy day, non-rainy day and quasi-rainy day, wherein dark sky, non-dark sky, rainy day and non-rainy day are conventional types, and quasi-dark sky and quasi-rainy day are characteristic types; An actuator unit is used to control the vehicle based on the discrimination type in combination with a preset actuator, the actuator unit including a body controller, body headlights and / or wipers, wherein when the discrimination type is the conventional type, the vehicle is controlled in direct response to the actuator; when the discrimination type is the characteristic type, the actuator is controlled based on the forward-looking sensor data.

9. The device for improving light rainfall sensitivity by multi-sensor fusion according to claim 8, characterized in that: The light and rain sensor data include light data and rain data, and the light and rain algorithm module includes a light perception algorithm and a rain perception algorithm, wherein the light perception algorithm processes the light data to obtain dark day, non-dark day and quasi-dark day results, and the rain perception algorithm processes the rain data to obtain rainy day, non-rainy day and quasi-rainy day. The image algorithm analyzes and processes the forward-looking sensor data to obtain image categories, wherein the image categories include non-rainy daytime, rainy daytime, non-rainy dark day, rainy dark day, tunnel / garage and unknown categories.

10. The device for improving light rainfall sensitivity by multi-sensor fusion according to claim 9, characterized in that: When the vehicle headlights are in the automatic gear, based on the output result of the light perception algorithm being dark, the control algorithm module controls the vehicle headlights to automatically turn on through the vehicle body controller; when the output result is non-dark, the control algorithm module controls the vehicle headlights not to turn on through the vehicle body controller; when the output result is quasi-dark, the forward-looking sensor data is analyzed based on the image algorithm, wherein if the analysis result is a tunnel / garage, the control algorithm module controls the vehicle headlights to turn on through the vehicle body controller, otherwise not to turn on; and when the wipers are in the automatic gear, based on the output result of the rain perception being rainy, the control algorithm module controls the wipers to automatically turn on through the vehicle body controller; when the output processing result is non-rainy, the control algorithm module controls the wipers not to turn on through the vehicle body controller; when the output result is quasi-rainy, the forward-looking sensor data is analyzed based on the image algorithm, wherein if the analysis result is rainy daytime or rainy dark, the control algorithm module controls the wipers to turn on through the vehicle body controller, otherwise not to turn on.

Images