Vehicle window defogging method and device, electronic device, and storage medium
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHONGQING CHANGAN TECH CO LTD
- Filing Date
- 2023-05-30
- Publication Date
- 2026-06-30
AI Technical Summary
When drivers manually adjust the car's air conditioning, they may use incorrect defogging methods, which can worsen the fogging and affect safe driving.
By acquiring image data of vehicle windows, a pre-trained Unet convolutional neural network model is used to detect fogging of vehicle windows. Based on vehicle driving data and driver and passenger information, the target defogging device and defogging strategy are automatically determined, and the defogging device is controlled to perform defogging operation.
It automatically detects window fogging and selects appropriate defogging equipment and strategies, avoiding user misoperation, improving vehicle driving safety and defogging efficiency, and meeting user preferences.
Smart Images
Figure CN116749914B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of automotive electronics technology, specifically to a method, apparatus, electronic device, and storage medium for defogging car windows. Background Technology
[0002] In cold weather, because the temperature difference between the inside and outside of the car is too large, the saturated vapor pressure on the surface of the car glass is lower than the vapor pressure inside the car. As a result, water vapor gathers on the glass surface and adheres to the car window glass in the form of tiny water droplets, causing the car window to fog up.
[0003] However, drivers currently tend to manually adjust the car's air conditioning based on their own experience, which can lead to incorrect adjustments and increased fogging, affecting safe driving. Summary of the Invention
[0004] In order to solve the above-mentioned technical problems, or at least partially solve the above-mentioned technical problems, this application provides a method, apparatus, electronic device and storage medium for defogging car windows.
[0005] Firstly, this application provides a method for defogging vehicle windows, including:
[0006] Acquire window image data captured from at least one vehicle window;
[0007] The window image data of each of the aforementioned windows is input into a neural network model so that the neural network model outputs the window fogging status corresponding to each of the aforementioned window image data;
[0008] Based on the fogging conditions of each vehicle window, a target defogging device for defogging the fogged windows is determined, and a defogging strategy is also determined.
[0009] The control target defogging device performs defogging on the fogged car window according to the defogging strategy.
[0010] Optionally, the neural network model includes: a cascaded input layer, a first convolutional layer, a first max pooling layer, a second convolutional layer, a second max pooling layer, a first deconvolutional layer, a second deconvolutional layer, a fully connected layer, and an activation function layer;
[0011] The input layer is used to input the window image data;
[0012] The first convolutional layer is used to extract image features from the window image data to obtain a first window feature vector;
[0013] The first max pooling layer is used to perform dimensionality reduction processing on the feature vector of the first window to obtain the dimensionality reduction vector of the first window.
[0014] The second convolutional layer is used to extract the image features of the first window's dimension reduction vector to obtain the second window's feature vector;
[0015] The second max pooling layer is used to reduce the dimensionality of the second window feature vector to obtain the second window dimensionality-reduced vector;
[0016] The first deconvolutional layer and the second deconvolutional layer are used to perform deconvolution processing on the second window dimension reduction vector to obtain a window dimension reduction feature vector with the same size as the window image data.
[0017] The fully connected layer is used to transform the two-dimensional window dimensionality reduction feature vector into a one-dimensional window dimensionality reduction feature vector.
[0018] The activation function layer is used to map the one-dimensional window dimensionality reduction feature vector to a preset data range.
[0019] Optionally, a target defogging device for defogging the fogged windows is determined based on the fogging condition of each window, and a defogging strategy is determined, including:
[0020] The fogged windows are identified based on the fogging conditions described for each window.
[0021] According to the preset correspondence between vehicle windows and defogging devices, the target defogging device for defogging the fogged vehicle windows is determined;
[0022] Obtain vehicle driving data;
[0023] A defogging strategy for defogging the fogged windows is determined based on the vehicle driving data.
[0024] Optionally, a defogging strategy for defogging the fogged window is determined based on the vehicle driving data, including:
[0025] If the vehicle's driving status is determined to be "driving" based on the vehicle driving data, the defogging strategy is determined to be the first defogging strategy. The first defogging strategy is used to instruct the target defogging device to defog by blowing cold air toward the fogged window at maximum airflow.
[0026] Alternatively, if the vehicle's driving status is determined to be "not driven out" based on the vehicle driving data, the defogging strategy is determined to be the second defogging strategy. The second defogging strategy is used to instruct the target defogging device to defog by blowing warm air towards the fogged window with a minimum airflow.
[0027] Optionally, a defogging strategy for defogging the fogged window is determined based on the vehicle driving data, including:
[0028] If the target driving distance of the vehicle is determined to be a short distance based on the vehicle driving data, the defogging strategy is determined to be the first defogging strategy. The first defogging strategy is used to instruct the target defogging device to defog by blowing cold air toward the fogged window at maximum airflow.
[0029] Alternatively, if the target driving distance of the vehicle is determined to be a long distance based on the vehicle driving data, the defogging strategy is determined to be a second defogging strategy. The second defogging strategy is used to instruct the target defogging device to defog by blowing warm air toward the fogged window with a minimum airflow.
[0030] Optionally, a defogging strategy for defogging the fogged window is determined based on the vehicle driving data, including:
[0031] If the current driving season is determined to be summer based on the vehicle driving data, the first defogging strategy is determined. The first defogging strategy is used to instruct the target defogging device to defog by blowing cold air toward the fogged window at maximum airflow.
[0032] Alternatively, if the current driving season is determined to be winter based on the vehicle driving data, the defogging strategy is determined to be the second defogging strategy. The second defogging strategy is used to instruct the target defogging device to defog by blowing warm air toward the fogged window with a minimum airflow.
[0033] Optionally, a defogging strategy for defogging the fogged window is determined based on the vehicle driving data, including:
[0034] Obtain driver and passenger information;
[0035] If the height information, seating information, and wind preference information of one or more drivers and passengers are determined based on the driver and passenger information;
[0036] A defogging strategy for defogging the fogged windows is determined based on the height information, seat information, airflow preference information, and vehicle driving data.
[0037] Secondly, this application provides a vehicle window defrosting device, comprising:
[0038] The acquisition module is used to acquire window image data captured from at least one window.
[0039] An input module is used to input the window image data of each window into a neural network model so that the neural network model outputs the window fogging status corresponding to each window image data.
[0040] The determination module is used to determine the target defogging device for defogging the fogged windows based on the fogging conditions of each window, and to determine the defogging strategy.
[0041] The control module is used to control the target defogging device to defog the fogged window according to the defogging strategy.
[0042] Thirdly, this application provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;
[0043] Memory, used to store computer programs;
[0044] The processor, when executing a program stored in memory, implements the window defogging method described in any of the first aspects.
[0045] Fourthly, this application provides a computer-readable storage medium storing a program for a method of defogging a vehicle window, wherein when the program of the method of defogging a vehicle window is executed by a processor, it implements the steps of the method of defogging a vehicle window as described in any of the first aspects.
[0046] The beneficial effects of this invention are:
[0047] This application embodiment determines the fogging condition of the vehicle window using a neural network model based on collected window image data. Then, it identifies the target defogging device and determines the defogging strategy based on the fogging condition. Finally, it controls the target defogging device to defog the fogged window according to the strategy. This achieves automatic detection of window fogging, automatically utilizing the appropriate defogging device and adopting the corresponding defogging strategy, eliminating the need for manual operation by the user. This simplifies user operation, improves vehicle driving safety, and, by automatically determining the defogging strategy, ensures the use of more accurate and efficient strategies. This avoids the problem of users mistakenly using incorrect strategies, which could worsen the fogging and affect safe driving. Attached Figure Description
[0048] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.
[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0050] Figure 1 A flowchart illustrating a method for defogging car windows provided in this application embodiment;
[0051] Figure 2 A structural diagram of a car window defroster provided in an embodiment of this application;
[0052] Figure 3 This is a structural diagram of an electronic device provided in an embodiment of this application. Detailed Implementation
[0053] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0054] Currently, drivers typically adjust the vehicle's air conditioning manually based on their experience, which can lead to incorrect adjustments and increased fogging, affecting safe driving. Therefore, this application provides a method, apparatus, electronic device, and storage medium for defogging vehicle windows.
[0055] This application provides a method for defogging car windows, such as... Figure 1 As shown, the vehicle defogging method includes:
[0056] Step S101: Acquire window image data captured from at least one window;
[0057] In this embodiment of the application, a vehicle window may refer to the windshield, the glass installed on the door, and the rear windshield, etc. The vehicle window image data may be collected by one or more cameras installed in the vehicle, and the cameras may collect the vehicle window image data of each window.
[0058] Step S102: Input the window image data of each window into the neural network model so that the neural network model outputs the window fogging status corresponding to each window image data;
[0059] In this embodiment of the application, the neural network model can be a pre-trained Unet convolutional neural network model, and the training method of the Unet convolutional neural network model can include the following steps:
[0060] First, images of car windows with and without fogging are obtained from the network to obtain the original dataset. Based on whether the images are fogged, the fogged images are labeled with label 1, and the non-foggy images are labeled with label 0, resulting in a dataset with labels indicating whether the images are fogged or not. The entire task can be regarded as a binary classification problem. For example, this invention can obtain 5000 photos as the training dataset for the model, and set the labels to 1 and 0 respectively. The original dataset is divided into training set, validation set and test set in a ratio of 6:2:2.
[0061] Secondly, the optimization algorithm for the Unet convolutional neural network model is set, and the Adam algorithm is used to optimize the model. Adam is a first-order optimization algorithm that can replace the traditional stochastic gradient descent process. It can iteratively update the neural network weights based on the training data. Adam designs independent adaptive learning rates Δθ for different parameters by calculating the first-order moment estimate and second-order moment estimate of the gradient. t .
[0062]
[0063] in Let represent the corrected first-order and second-order matrices. ∈ , and η are parameters that need to be adjusted during training.
[0064] For example, in this application embodiment, the learning rate can be 0.0001.
[0065] In this embodiment, the cross-entropy loss function is used to display the difference between the predicted and true values. Model parameters are pre-set, such as learning rate, kernel size, pooling layers, etc. The convolution kernel represents the visual perception domain, and the kernel size represents the resolution of that domain. A smaller kernel yields more obvious features from the image, leading to better image understanding and more image feature descriptions. Therefore, a 3×3 kernel is used with a stride of 1. Max pooling is used, selecting the maximum value of the image region as the pooled value for that region.
[0066] Since the entire task in this application is a binary classification problem, this application uses a fully connected layer with only one neuron and employs the sigmoid activation function to obtain the final output. The sigmoid formula is as follows:
[0067]
[0068] Where x is the input to the activation function layer.
[0069] Next, the Unet convolutional neural network model is trained. The images in the training set are input into the input layer, and then into the first convolutional layer (C1) with k1 kernels for convolution. Then, a pooling operation is performed through a max pooling layer (p1). The convolution and pooling are iterated twice to obtain the output k2 of the two-dimensional feature vector. Then, two upsampling (deconvolution) operations are performed to obtain a two-dimensional embedding vector of the same size as the input image. Next, two fully connected layers map it to a unique output. Finally, the sigmoid function is used to map the output to the range (0, 1) to obtain the output of the Unet convolutional neural network model, i.e., the predicted value.
[0070] Finally, the difference between the predicted value and the true value is calculated using the loss function. If the difference between the predicted value and the true value exceeds a preset threshold, the model parameters are adjusted until the model converges.
[0071] Afterwards, the model can be validated using a validation set and tested using a test set until a trained neural network model is obtained. Once the neural network model is trained, it can automatically output the window fogging status corresponding to the input window image data. The window fogging status can include two states: window fogged or window not fogged.
[0072] Step S103: Determine the target defogging device for defogging the fogged windows based on the fogging conditions of each window, and determine the defogging strategy;
[0073] After obtaining the fogging status of each car window, since the location of each window is different, the corresponding target defogging equipment is also different. Different vehicle driving data may also lead to different defogging strategies. Therefore, the target defogging equipment for defogging the fogged windows can be determined based on whether each window is fogged, and the defogging strategy can be determined.
[0074] For example, when it is determined that the windshield is fogged up, the target defogging device can be defogging device 1, and the defogging strategy can be to blow cold air towards the fogged window using the maximum airflow.
[0075] Step S104: Control the target defogging device to defog the fogged window according to the defogging strategy.
[0076] After determining the target defogging device and the defogging strategy, a control command can be sent to the target defogging device. The control command is used to instruct the target defogging device to defog the fogged window according to the defogging strategy.
[0077] This application embodiment determines the fogging condition of the vehicle window using a neural network model based on collected window image data. Then, it identifies the target defogging device and determines the defogging strategy based on the fogging condition. Finally, it controls the target defogging device to defog the fogged window according to the strategy. This achieves automatic detection of window fogging, automatically utilizing the appropriate defogging device and adopting the corresponding defogging strategy, eliminating the need for manual operation by the user. This simplifies user operation, improves vehicle driving safety, and, by automatically determining the defogging strategy, ensures the use of more accurate and efficient strategies. This avoids the problem of users mistakenly using incorrect strategies, which could worsen the fogging and affect safe driving.
[0078] In another embodiment of this application, the neural network model includes: a cascaded input layer, a first convolutional layer, a first max pooling layer, a second convolutional layer, a second max pooling layer, a first deconvolutional layer, a second deconvolutional layer, a fully connected layer, and an activation function layer;
[0079] The input layer is used to input the window image data;
[0080] The first convolutional layer is used to extract image features from the window image data to obtain a first window feature vector;
[0081] The first max pooling layer is used to perform dimensionality reduction processing on the feature vector of the first window to obtain the dimensionality reduction vector of the first window.
[0082] The second convolutional layer is used to extract the image features of the first window's dimension reduction vector to obtain the second window's feature vector;
[0083] The second max pooling layer is used to reduce the dimensionality of the second window feature vector to obtain the second window dimensionality-reduced vector;
[0084] The first deconvolutional layer and the second deconvolutional layer are used to perform deconvolution processing on the second window dimension reduction vector to obtain a window dimension reduction feature vector with the same size as the window image data.
[0085] The fully connected layer is used to transform the two-dimensional window dimensionality reduction feature vector into a one-dimensional window dimensionality reduction feature vector.
[0086] The activation function layer is used to map the one-dimensional window dimensionality reduction feature vector to a preset data range.
[0087] In this embodiment, features can be extracted and dimensionality reduced by passing through convolutional layers and max pooling layers twice. This facilitates the full extraction of advantageous features from the window image data. Furthermore, the deconvolution process through two deconvolutional layers can ultimately yield a window dimensionality-reduced feature vector of the same size as the window image data, which facilitates dimensionality transformation in subsequent fully connected layers and improves the accuracy of the model.
[0088] In another embodiment of this application, step S103 determines a target defogging device for defogging the fogged windows based on the fogging conditions of each window, and determines a defogging strategy, including:
[0089] Step S201: Determine the fogged windows based on the fogging condition of each window;
[0090] In practical applications, at least one car window may have some windows fogged up while others are not. Therefore, the fogging status of each car window can be used to determine whether each window is fogged up, and thus identify the fogged windows.
[0091] Step S202: According to the preset correspondence between vehicle windows and defogging devices, determine the target defogging device for defogging the fogged vehicle windows;
[0092] Each defogging device can be used to defog at least one vehicle window, and a correspondence between vehicle windows and defogging devices can be pre-established. In this step, a vehicle window that matches the fogged window can be found in the correspondence, and the defogging device corresponding to that window can be identified as the target defogging device.
[0093] When the same defrost device is used for at least two vehicle windows, a defrost priority can be set for each of the at least two windows. The defrost strategy can be to first adjust the tilt angle of the defrost device to face the window with the higher defrost priority, and then defog the window with the higher defrost priority first. After a period of time (the length of this period of time can be determined according to the time required for defogging the window with the higher priority), the defrost device is adjusted to face the window with the lower defrost priority, and then the window with the lower defrost priority is defog first. This allows the same defrost device to be used in a time-sharing manner according to priority, improving the efficiency of the defrost device and ensuring the effect of use.
[0094] Step S203: Obtain vehicle driving data;
[0095] In this embodiment of the application, the vehicle driving data includes: driving status, target driving distance, and current driving season, etc.
[0096] Step S204: Determine a defogging strategy for defogging the fogged windows based on the vehicle driving data.
[0097] In one embodiment of this application, if the vehicle's driving state is determined to be in motion based on the vehicle driving data, the defogging strategy is determined to be a first defogging strategy. The first defogging strategy is used to instruct the target defogging device to defog by blowing cold air toward the fogged window at maximum airflow.
[0098] Alternatively, if the vehicle's driving status is determined to be "not driven out" based on the vehicle driving data, the defogging strategy is determined to be the second defogging strategy. The second defogging strategy is used to instruct the target defogging device to defog by blowing warm air towards the fogged window with a minimum airflow.
[0099] In another embodiment of this application, if the target driving distance of the vehicle is determined to be a short distance based on the vehicle driving data, the defogging strategy is determined to be a first defogging strategy. The first defogging strategy is used to instruct the target defogging device to defog by blowing cold air toward the fogged window at maximum airflow.
[0100] Alternatively, if the target driving distance of the vehicle is determined to be a long distance based on the vehicle driving data, the defogging strategy is determined to be a second defogging strategy. The second defogging strategy is used to instruct the target defogging device to defog by blowing warm air toward the fogged window with a minimum airflow.
[0101] In another embodiment of this application, if the current driving season is determined to be summer based on the vehicle driving data, the defogging strategy is determined to be the first defogging strategy. The first defogging strategy is used to instruct the target defogging device to defog by blowing cold air toward the fogged window at maximum airflow.
[0102] Alternatively, if the current driving season is determined to be winter based on the vehicle driving data, the defogging strategy is determined to be the second defogging strategy. The second defogging strategy is used to instruct the target defogging device to defog by blowing warm air toward the fogged window with a minimum airflow.
[0103] The embodiments of this application can automatically determine the target defogging device and defogging strategy, eliminating the need for drivers and passengers to make their own selections. This avoids situations where drivers and passengers select the wrong target defogging device and defogging strategy, resulting in poor defogging effects or even more severe fogging. It improves the efficiency of window defogging and makes it easier for users to operate and use.
[0104] In another embodiment of this application, a defogging strategy for defogging the fogged window is determined based on the vehicle driving data, including:
[0105] Step S301: Obtain driver and passenger information;
[0106] In practical applications, different drivers and passengers may have different heights and different preferences for airflow. In this embodiment, the driver and passenger information may include: height information, seat information, and airflow preference information. Height information is used to indicate the height of the driver and passenger, and airflow preference information is used to indicate which airflow angle the driver and passenger prefers when the air conditioner is blowing cold or warm air, such as: avoiding face airflow, preferring face airflow, avoiding shoulder airflow, etc.
[0107] Step S302: If the height information, seat information and air blowing preference information of one or more drivers and passengers are determined based on the driver and passenger information;
[0108] In this step, the height, seating information, and airflow preference information of each driver and passenger can be directly extracted from the driver and passenger information.
[0109] Step S303: Determine a defogging strategy for defogging the fogged windows based on the height information, seat information, airflow preference information, and vehicle driving data.
[0110] In this step, the target body part that is preferred to be blown or avoided can be determined first based on the blowing preference information. The target coordinate position of the preferred body part or the target body part that is avoided can be calculated based on the height and seating information. The angle of the reflected wind can be determined based on the correspondence between the blowing angle and the reflected wind angle of different defogging devices. The allowable blowing angle is determined so that the reflected wind blows towards or avoids blowing towards the target coordinate position.
[0111] For example, if a user A is sensitive to wind, to prevent the wind reflected from the car window from blowing onto the user's face, we can first determine the target coordinates of the user A's face based on the user A's height and seating information, and then calculate the allowable wind angle so that the reflected wind can avoid the target coordinates.
[0112] For example, if a user B likes to be exposed to the wind, then to make the wind reflected from the car window blow towards the user's face, we can first determine the target coordinates of the user's face based on the user's height and seating information, and then calculate the allowable wind angle so that the reflected wind blows towards the target coordinates.
[0113] For example, if a user C is sensitive to wind in their shoulders, then to make the wind reflected from the car window blow towards the user C's shoulders, we can first determine the target coordinates of the user C's shoulders based on the user C's height and seating information, and then calculate the allowable wind angle to blow towards the target coordinates.
[0114] In conjunction with the aforementioned embodiments, based on the preliminary defogging strategy determined from the vehicle driving data, the maximum airflow is used to blow cold air toward the fogged window at the permissible airflow angle, or the minimum airflow is used to blow warm air toward the fogged window.
[0115] This application embodiment can automatically determine the defogging strategy based on driver and passenger information and vehicle driving data, without requiring the driver and passengers to make their own selections. This avoids situations where the driver and passengers choose the wrong defogging strategy, resulting in poor defogging effect or even more severe fogging. It improves the efficiency of window defogging, meets the user's airflow preferences, and is convenient for user operation and use.
[0116] In another embodiment of this application, a vehicle window defrosting device is also provided, such as... Figure 2 As shown, it includes:
[0117] The acquisition module 11 is used to acquire window image data collected from at least one window.
[0118] Input module 12 is used to input the window image data of each window into the neural network model so that the neural network model outputs the window fogging status corresponding to each window image data;
[0119] The determining module 13 is used to determine the target defogging device for defogging the fogged windows based on the fogging conditions of each window, and to determine the defogging strategy.
[0120] Control module 14 is used to control the target defogging device to defog the fogged window according to the defogging strategy.
[0121] In another embodiment of this application, an electronic device is also provided, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus.
[0122] Memory, used to store computer programs;
[0123] The processor, when executing a program stored in memory, implements the window defogging method described in any of the foregoing method embodiments.
[0124] The electronic device provided in this invention uses a processor to execute a program stored in its memory. Based on collected window image data, the processor uses a neural network model to determine the fogging condition of the windows, then identifies a target defogging device and determines a defogging strategy. Finally, it controls the target defogging device to defog the fogged windows according to the strategy. This achieves automatic detection of window fogging, automatic use of appropriate defogging devices and strategies, eliminating the need for manual user operation, improving user convenience and vehicle safety. Furthermore, by automatically determining the defogging strategy, more accurate and efficient strategies can be used automatically, preventing users from using incorrect strategies that could worsen the fogging and affect safe driving.
[0125] The communication bus 1140 mentioned in the above-mentioned electronic device can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus 1140 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.
[0126] The communication interface 1120 is used for communication between the above-mentioned electronic device and other devices.
[0127] The memory 1130 may include random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.
[0128] The processor 1110 mentioned above can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.
[0129] In another embodiment of this application, a computer-readable storage medium is provided, on which a program for a vehicle window defogging method is stored. When the program for the vehicle window defogging method is executed by a processor, it implements the steps of the vehicle window defogging method described in any of the foregoing method embodiments.
[0130] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0131] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.
Claims
1. A method for defrosting car windows, characterized in that, include: Acquire window image data captured from at least one vehicle window; The window image data of each of the aforementioned windows is input into a neural network model so that the neural network model outputs the window fogging status corresponding to each of the aforementioned window image data; Based on the fogging conditions of each vehicle window, a target defogging device for defogging the fogged windows is determined, and a defogging strategy is also determined. Based on the fogging conditions of each vehicle window, a target defogging device for defogging the fogged windows is determined, and a defogging strategy is determined, including: The fogged windows are identified based on the fogging conditions described for each window. According to the preset correspondence between car windows and defogging devices, a target defogging device is determined for defogging the fogged car windows; when the same defogging device corresponds to at least two car windows, a defogging priority is set for each of the at least two car windows. The defogging strategy adopted is to first adjust the tilt angle of the defogging device so that the defogging device faces the car window with the high defogging priority, so as to defog the car window with the high defogging priority first; after a preset time, the defogging device is adjusted to face the car window with the low defogging priority, so as to defog the car window with the low defogging priority first. Acquire vehicle driving data, including: driving status, target driving distance, and current driving season; A defogging strategy for defogging the fogged windows is determined based on the vehicle driving data. Based on the vehicle driving data, a defogging strategy is determined for defogging the fogged windows, including: If the vehicle's driving status is determined to be in motion based on the vehicle driving data, a preliminary defogging strategy of blowing cold air towards the fogged window with the maximum airflow is determined; or, if the vehicle's driving status is determined to be not yet moved based on the vehicle driving data, a preliminary defogging strategy of blowing warm air towards the fogged window with the minimum airflow is determined. If the target driving distance of the vehicle is determined to be a short distance based on the vehicle driving data, a preliminary defogging strategy of blowing cold air towards the fogged window with the maximum airflow is determined; or, if the target driving distance of the vehicle is determined to be a long distance based on the vehicle driving data, a preliminary defogging strategy of blowing warm air towards the fogged window with the minimum airflow is determined. If the current driving season is determined to be summer based on the vehicle driving data, a preliminary defogging strategy of blowing cold air towards the fogged window with the maximum airflow is determined; or, if the current driving season is determined to be winter based on the vehicle driving data, a preliminary defogging strategy of blowing warm air towards the fogged window with the minimum airflow is determined. Obtain driver and passenger information, and determine the permissible airflow angle based on the driver and passenger information; Based on the vehicle driving data, the preliminary defogging strategy is determined by blowing cold air towards the fogged window at the maximum airflow at the allowed airflow angle, or blowing warm air towards the fogged window at the minimum airflow. The control target defogging device performs defogging on the fogged car window according to the defogging strategy.
2. The method for defrosting car windows according to claim 1, characterized in that, The neural network model includes: a cascaded input layer, a first convolutional layer, a first max pooling layer, a second convolutional layer, a second max pooling layer, a first deconvolutional layer, a second deconvolutional layer, a fully connected layer, and an activation function layer; The input layer is used to input the window image data; The first convolutional layer is used to extract image features from the window image data to obtain a first window feature vector; The first max pooling layer is used to perform dimensionality reduction processing on the feature vector of the first window to obtain the dimensionality reduction vector of the first window. The second convolutional layer is used to extract the image features of the first window's dimension reduction vector to obtain the second window's feature vector; The second max pooling layer is used to reduce the dimensionality of the second window feature vector to obtain the second window dimensionality-reduced vector; The first deconvolutional layer and the second deconvolutional layer are used to perform deconvolution processing on the second window dimension reduction vector to obtain a window dimension reduction feature vector with the same size as the window image data. The fully connected layer is used to transform the two-dimensional window dimensionality reduction feature vector into a one-dimensional window dimensionality reduction feature vector. The activation function layer is used to map the one-dimensional window dimensionality reduction feature vector to a preset data range.
3. The method for defrosting car windows according to claim 1, characterized in that, Based on the vehicle driving data, a defogging strategy is determined for defogging the fogged windows, including: If the vehicle's driving status is determined to be "driving" based on the vehicle driving data, the defogging strategy is determined to be the first defogging strategy. The first defogging strategy is used to instruct the target defogging device to defog by blowing cold air toward the fogged window at maximum airflow. Alternatively, if the vehicle's driving status is determined to be "not driven out" based on the vehicle driving data, the defogging strategy is determined to be the second defogging strategy. The second defogging strategy is used to instruct the target defogging device to defog by blowing warm air towards the fogged window with a minimum airflow.
4. The method for defrosting car windows according to claim 1, characterized in that, Based on the vehicle driving data, a defogging strategy is determined for defogging the fogged windows, including: If the target driving distance of the vehicle is determined to be a short distance based on the vehicle driving data, the defogging strategy is determined to be the first defogging strategy. The first defogging strategy is used to instruct the target defogging device to defog by blowing cold air toward the fogged window at maximum airflow. Alternatively, if the target driving distance of the vehicle is determined to be a long distance based on the vehicle driving data, the defogging strategy is determined to be a second defogging strategy. The second defogging strategy is used to instruct the target defogging device to defog by blowing warm air toward the fogged window with a minimum airflow.
5. The method for defrosting car windows according to claim 1, characterized in that, Based on the vehicle driving data, a defogging strategy is determined for defogging the fogged windows, including: If the current driving season is determined to be summer based on the vehicle driving data, the first defogging strategy is determined. The first defogging strategy is used to instruct the target defogging device to defog by blowing cold air toward the fogged window at maximum airflow. Alternatively, if the current driving season is determined to be winter based on the vehicle driving data, the defogging strategy is determined to be the second defogging strategy. The second defogging strategy is used to instruct the target defogging device to defog by blowing warm air toward the fogged window with a minimum airflow.
6. The method for defrosting car windows according to claim 1, characterized in that, Based on the vehicle driving data, a defogging strategy is determined for defogging the fogged windows, including: Obtain driver and passenger information; If the height information, seating information, and wind preference information of one or more drivers and passengers are determined based on the driver and passenger information; A defogging strategy for defogging the fogged windows is determined based on the height information, seat information, airflow preference information, and vehicle driving data.
7. A vehicle window defroster, characterized in that, include: The acquisition module is used to acquire window image data captured from at least one window. An input module is used to input the window image data of each window into a neural network model so that the neural network model outputs the window fogging status corresponding to each window image data. The determination module is used to determine the target defogging device for defogging the fogged windows based on the fogging conditions of each window, and to determine the defogging strategy. Based on the fogging condition of each vehicle window, a target defogging device is determined for defogging the fogged windows, and a defogging strategy is determined, including: identifying the fogged windows based on the fogging condition of each vehicle window; determining the target defogging device for defogging the fogged windows according to a preset correspondence between vehicle windows and defogging devices; when the same defogging device corresponds to at least two vehicle windows, setting a defogging priority for each of the at least two vehicle windows, and adopting a defogging strategy that first adjusts the tilt angle of the defogging device so that the defogging device faces the window with the higher defogging priority, so as to defog the window with the higher defogging priority. Prioritize defogging. After a preset time, adjust the defogging device towards windows with low defogging priority to prioritize defogging those windows. Acquire vehicle driving data, including driving status, target driving distance, and current driving season. Determine a defogging strategy for the fogged windows based on the vehicle driving data. This strategy includes: if the vehicle's driving status is determined to be in motion based on the driving data, determine to use maximum airflow towards the fogged windows. The initial defogging strategy is to blow cold air into the window; or, if the vehicle's driving status is determined to be not yet moved based on the vehicle driving data, a preliminary defogging strategy is determined to be to blow warm air into the fogged window using the minimum airflow; if the target driving distance of the vehicle is determined to be a short distance based on the vehicle driving data, a preliminary defogging strategy is determined to be to blow cold air into the fogged window using the maximum airflow; or, if the target driving distance of the vehicle is determined to be a long distance based on the vehicle driving data, a preliminary defogging strategy is determined to be to blow warm air into the fogged window using the minimum airflow; if the target driving distance is determined to be a long distance based on the vehicle driving data... If the current driving season is determined to be summer, a preliminary defogging strategy is determined to use the maximum airflow to blow cold air towards the fogged window; or, if the current driving season is determined to be winter based on the vehicle driving data, a preliminary defogging strategy is determined to use the minimum airflow to blow warm air towards the fogged window; driver and passenger information is obtained, and the permissible airflow angle is determined based on the driver and passenger information; based on the preliminary defogging strategy determined by the vehicle driving data, the strategy of blowing cold air towards the fogged window with the maximum airflow at the permissible airflow angle, or blowing warm air towards the fogged window with the minimum airflow, is determined as the defogging strategy. The control module is used to control the target defogging device to defog the fogged window according to the defogging strategy.
8. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; The processor, when executing a program stored in the memory, implements the window defogging method according to any one of claims 1 to 6.
9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program for a method of defogging a vehicle window, which, when executed by a processor, implements the steps of the method of defogging a vehicle window as described in any one of claims 1-6.