A method, system, device and storage medium for preventing car windows from being pinched.
By collecting environmental parameters and combining them with a temporal convolutional network and attention mechanism model to calculate the entry probability and intent confidence, the problem of misjudgment in the existing window anti-pinch control method under environmental interference is solved, achieving higher recognition accuracy and reliability.
Patent Information
- Application Number
- CN202611370679.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-04
- Publication Date
- 2026-11-17
AI Technical Summary
Existing methods for preventing car windows from being pinched fail to effectively combine environmental parameters such as lighting to verify image recognition results, which can easily lead to misjudgments in environments such as backlight, shadow, or low light at night, making it impossible to generate reliable and accurate anti-pinch control strategies for car windows.
By synchronously collecting environmental parameters after the window is raised, and combining temporal convolutional networks and attention mechanism models, the probability of human body parts entering the dangerous area and the confidence level of intent are calculated. The window anti-pinch control scheme is then comprehensively decided to distinguish between accidental occlusion and active insertion behavior.
It effectively reduces the probability of false triggering and missed detection of the window anti-pinch function, improves the accuracy and reliability of window lifting control, and adapts to the recognition accuracy under different lighting and driving speed conditions.
Smart Images

Figure CN123109496A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the technical field of anti-pinch control for automotive windows, and in particular to a method, system, device, and storage medium for anti-pinch control of automotive windows. Background Technology
[0002] Anti-pinch windows, as a safety feature, have become standard equipment in automobiles. However, most existing anti-pinch control methods rely solely on image recognition results to output control commands, without combining these results with environmental parameters such as lighting conditions for verification. This makes them susceptible to inaccurate recognition of human body movements due to environmental interference such as backlighting, shadows, or low-light conditions at night, leading to false alarms and an inability to generate reliable and accurate anti-pinch control strategies. Summary of the Invention
[0003] This application aims to at least address the technical problems existing in the prior art. To this end, this application proposes a method, system, device, and storage medium for controlling anti-pinch mechanisms in automobile windows, which can reduce the probability of false triggering and missed detection of anti-pinch mechanisms, and improve the accuracy and reliability of window lifting control.
[0004] The first aspect of this application provides a method for controlling the anti-pinch function of automotive windows, comprising the following steps: Upon receiving notification that any window of the vehicle is raised, obtain environmental parameter values; Based on the environmental parameter values, a sequence of in-vehicle images is determined; Based on the in-vehicle image sequence, the entry probability and intent confidence are determined, wherein the entry probability is used to characterize the probability value of a preset human body part extending into the preset dangerous area of the vehicle window, and the intent confidence is used to characterize the confidence level of the preset human body part extending into the vehicle. Based on the environmental parameter values, the entry probability, and the intent confidence level, a vehicle window anti-pinch control scheme is determined.
[0005] The vehicle window anti-pinch control method according to the embodiments of this application has at least the following beneficial effects: This application collects environmental parameters simultaneously after detecting a window raising command, and identifies corresponding in-vehicle image sequences based on these parameters. It calculates the entry probability and behavioral intent confidence of human body parts in dangerous areas, and then combines environmental parameters, entry probability, and intent confidence to jointly determine the window anti-pinch control scheme. This effectively avoids image recognition deviations caused by changes in lighting, avoids misidentification problems caused by single visual judgment, distinguishes between accidental occlusion and genuine active insertion behavior, significantly reduces the probability of false triggering and missed detection of window anti-pinch, and improves the accuracy and reliability of window raising and lowering control.
[0006] A second aspect of this application provides an anti-pinch control system for automotive windows, the automotive window anti-pinch control system comprising: The data acquisition module is used to acquire environmental parameter values when receiving information that any window is raised. The in-vehicle image sequence determination module is used to determine the in-vehicle image sequence based on the environmental parameter values; An entry probability and intent confidence determination module is used to determine the entry probability and intent confidence based on the in-vehicle image sequence, wherein the entry probability is used to characterize the probability value of a preset human body part extending into a preset dangerous area of the vehicle window, and the intent confidence is used to characterize the confidence of the preset human body part extending into the vehicle. The vehicle window anti-pinch control scheme determination module is used to determine the vehicle window anti-pinch control scheme based on the environmental parameter values, the entry probability, and the intent confidence level.
[0007] This system collects environmental parameters simultaneously upon detecting a window raising command, and then matches the corresponding in-vehicle image sequence with these parameters for identification. It calculates the entry probability and behavioral intent confidence of human body parts in dangerous areas, and then combines environmental parameters, entry probability, and intent confidence to jointly determine the window anti-pinch control scheme. This effectively avoids image recognition deviations caused by changes in lighting, avoids misidentification problems caused by single visual judgment, distinguishes between accidental occlusion and genuine active insertion behavior, significantly reduces the probability of false triggering and missed detection of window anti-pinch, and improves the accuracy and reliability of window raising and lowering control.
[0008] A third aspect of this application provides an electronic device including at least one processor and a memory for communicatively connecting to the processor; the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to cause the at least one processor to perform an anti-pinch control method for automotive windows as described in the first aspect of this application.
[0009] A fourth aspect of this application provides a computer-readable storage medium storing computer-executable instructions for causing a computer to perform an anti-pinch control method for automobile windows as described in the first aspect of this application.
[0010] It should be noted that the beneficial effects of the third and fourth aspects of this application compared with the prior art are the same as the beneficial effects of the above-mentioned anti-pinch control method for automobile windows compared with the prior art, and will not be described in detail here.
[0011] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description
[0012] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a flowchart illustrating an embodiment of the anti-pinch control method for automobile windows provided in this application; Figure 2 This is a schematic diagram of an embodiment of the anti-pinch control system for automobile windows provided in this application; Figure 3 This is a schematic diagram of the structure of an embodiment of the electronic device provided in this application. Detailed Implementation
[0013] The embodiments of this application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application.
[0014] In the description of this application, the use of terms such as "first," "second," etc., is for the purpose of distinguishing technical features only and should not be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated or the order of the technical features indicated.
[0015] In the description of this application, it should be understood that the orientation descriptions, such as up, down, etc., are based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.
[0016] In the description of this application, it should be noted that, unless otherwise explicitly defined, terms such as "setup," "installation," and "connection" should be interpreted broadly, and those skilled in the art can reasonably determine the specific meaning of the above terms in this application in conjunction with the specific content of the technical solution.
[0017] Anti-pinch windows, as a safety feature, have become standard equipment in automobiles. However, most existing anti-pinch control methods rely solely on image recognition results to output control commands, without combining these results with environmental parameters such as lighting conditions for verification. This makes them susceptible to inaccurate recognition of human body movements due to environmental interference such as backlighting, shadows, or low-light conditions at night, leading to false alarms and an inability to generate reliable and accurate anti-pinch control strategies.
[0018] To address the aforementioned technical deficiencies, this application provides a method, system, device, and storage medium for controlling anti-pinch features of automotive windows.
[0019] Please see Figure 1 This is a flowchart illustrating an anti-pinch control method for automotive windows provided in an embodiment of this application. The method is applied to an electronic device, which may be a server, etc. Figure 1 As shown, the anti-pinch control method for automobile windows includes: Step S101: Upon receiving information that any vehicle window is raised, obtain environmental parameter values; The environmental parameter values mentioned above may include vehicle speed and light intensity, and may also include the vehicle interior temperature.
[0020] In step S101, when receiving information that any window is raised, the environmental parameter values can be obtained by real-time collection of feedback signals from the window lift switches and window drive motors of each window in the vehicle body control module (BCM), and continuous monitoring of the hardware trigger status of the four windows. When the window raise trigger signal of any one of the driver's side, passenger side, left rear, or right rear windows changes from low level to high level, it is determined that information that any window is raised has been received. The interior temperature is collected by the installed interior temperature sensor, the light intensity value is collected by the installed light sensor, and the vehicle speed is collected by the installed wheel speed sensor.
[0021] Step S102: Determine the in-vehicle image sequence based on environmental parameter values; Step S103: Based on the in-vehicle image sequence, determine the entry probability and intent confidence, wherein the entry probability is used to characterize the probability value of a preset human body part extending into the preset dangerous area of the vehicle window, and the intent confidence is used to characterize the confidence of the preset human body part extending into the vehicle. The aforementioned preset dangerous areas for vehicle windows can be pre-set values according to actual needs.
[0022] The aforementioned preset human body parts can be values pre-set according to actual needs, such as the person's hands or head.
[0023] Step S104: Based on environmental parameter values, entry probability, and intent confidence, determine the anti-pinch control scheme for car windows.
[0024] This application collects environmental parameters simultaneously after detecting a window raising command, and identifies corresponding in-vehicle image sequences based on these parameters. It calculates the entry probability and behavioral intent confidence of human body parts in dangerous areas, and then combines environmental parameters, entry probability, and intent confidence to jointly determine the window anti-pinch control scheme. This effectively avoids image recognition deviations caused by changes in lighting, avoids misidentification problems caused by single visual judgment, distinguishes between accidental occlusion and genuine active insertion behavior, significantly reduces the probability of false triggering and missed detection of window anti-pinch, and improves the accuracy and reliability of window raising and lowering control.
[0025] In some embodiments, step S103 may include steps S201 to S202: Step S201: Extract motion trajectory features of in-vehicle image sequences using a temporal convolutional network model. The motion trajectory features include motion direction vector, motion speed, motion acceleration, and distance from the boundary between the preset human body part and the preset dangerous area of the vehicle window. Step S202: Input the motion trajectory features into the attention-based intention prediction model to obtain the entry probability and intention confidence output by the attention-based intention prediction model.
[0026] The aforementioned attention-based intent prediction model includes a spatial attention mechanism, a channel attention mechanism, and a dual-branch output layer. The spatial attention mechanism identifies the relative positional relationship between a preset human body part and a preset dangerous area of the car window. The channel attention mechanism adaptively weights the importance of different feature channels. The probability of a preset human body part extending into the preset dangerous area of the car window is calculated using the sigmoid activation function of the probability branch of the dual-branch output layer. The intent confidence is calculated by combining the confidence branch of the dual-branch output layer with the dispersion of the attention weight distribution.
[0027] This application extracts multi-dimensional trajectory features, including motion direction, velocity, acceleration, and region boundary distance, from image sequences using a temporal convolutional network. This fully captures continuous dynamic changes in human body parts. Furthermore, by leveraging an attention mechanism, key motion features are given higher weights, accurately outputting entry probability and intent confidence. This effectively filters interference from static object occlusion, distinguishes between random occlusion and active reaching actions, significantly improves behavior recognition accuracy, and provides reliable behavioral quantification indicators for anti-pinch decisions on car windows.
[0028] In some embodiments, step S104 may include steps S301 to S302: Step S301: When the light intensity value is less than the preset light intensity threshold and the vehicle speed is less than the preset vehicle speed threshold, a first warning level is determined based on the first preset probability threshold, the second preset probability threshold, the entry probability, the first preset confidence threshold, the second preset confidence value, and the intent confidence. The first preset probability threshold is less than the second preset probability threshold, and the first preset confidence threshold is less than the second preset confidence value. The aforementioned preset light intensity threshold can be a value that is preset according to actual needs.
[0029] The aforementioned preset vehicle speed threshold can be a value that is pre-set according to actual needs.
[0030] The aforementioned first preset probability threshold can be a value preset according to actual needs.
[0031] The aforementioned second preset probability threshold can be a value preset according to actual needs.
[0032] The aforementioned first preset reliability threshold can be a value preset according to actual needs.
[0033] The aforementioned second preset reliability value can be a value preset according to actual needs.
[0034] The aforementioned first warning level can be a false alarm level, a warning level, an intervention level, or a trap level.
[0035] Step S302: Based on the first warning level and the preset level response strategy, determine the anti-pinch control scheme for the car window.
[0036] The aforementioned preset response strategies may include Scheme 1, Scheme 2, Scheme 3, and Scheme 4. Among them, Scheme 1 may not trigger a warning and allow normal window closing; Scheme 2 may prompt the user "Do not stick your hands out of the window" through voice and display screen, while reducing the window closing speed to 50% of the normal speed preset according to actual needs; Scheme 3 may pause the window closing action, maintain the current position, and continue voice prompts; Scheme 4 may immediately implement anti-pinch measures and control the window to close.
[0037] In step S302, the above-mentioned anti-pinch control scheme for the car window based on the warning level and the preset level response strategy can be as follows: if the first warning level is a false alarm level, the first preset level response strategy scheme can be used as the anti-pinch control scheme for the car window; if the first warning level is a warning level, the second preset level response strategy can be used as the anti-pinch control scheme for the car window; if the first warning level is an intervention level, the third preset level response strategy can be used as the anti-pinch control scheme for the car window; and if the first warning level is an anti-pinch level, the fourth preset level response strategy can be used as the anti-pinch control scheme for the car window.
[0038] This application addresses high-interference conditions such as low light and low speed by setting two progressive probability thresholds and confidence thresholds. It combines entry probability and intent confidence to divide into multiple warning levels and matches corresponding response strategies. This allows for graded identification of dangerous behaviors, effectively reducing misjudgments caused by image recognition in low-light environments, differentiating the strength of risks, and enabling graded responses to window anti-pinch actions. This avoids meaningless frequent reversals and provides timely braking for high-intrusion risks, significantly reducing the probability of false triggering and missed detection of window anti-pinch mechanisms, and improving the accuracy and reliability of window lifting control.
[0039] In some embodiments, step S301 may include steps S401 to S404: Step S401: If the probability of entering is less than or equal to the first preset probability threshold, or if the confidence level of intent is less than or equal to the first preset confidence threshold, the false alarm level shall be used as the first warning level. Step S402: When the entry probability is greater than the first preset probability threshold and less than the second preset probability threshold, and the intention confidence is greater than the first preset confidence threshold, the warning level is set as the first warning level; when the entry probability is greater than the first preset probability threshold, and the intention confidence is greater than the first preset confidence threshold and less than the second preset confidence value, the warning level is set as the first warning level. Step S403: When the probability of entry is greater than or equal to the second preset probability threshold and less than one, and the confidence level of intent is greater than or equal to the second preset confidence value, the intervention level is set as the first warning level. Step S404: When the probability of entry is equal to one, the anti-pinch level is set as the first warning level.
[0040] This application divides the entry probability and intent confidence into intervals by setting two probability thresholds and two confidence thresholds, distinguishing four risk levels: false alarm, warning, intervention, and anti-pinch. This allows for a refined quantification and classification of the danger level of human insertion behavior, effectively filtering out invalid actions caused by minor misidentifications. It matches corresponding handling logic to different levels of danger, avoiding the inconvenience caused by frequent window opening and stopping, and promptly implementing anti-pinch protection for high-risk intrusion behaviors, thereby improving the reliability of window lifting control.
[0041] In some embodiments, step S104 may include steps S501 to S502: Step S501: When the vehicle speed is greater than or equal to a preset vehicle speed threshold, a second warning level is determined based on a third preset probability threshold, an entry probability, a third preset confidence threshold, and an intent confidence level, wherein the third preset probability threshold is less than the first preset probability threshold, and the third preset confidence threshold is less than the first preset confidence value. The aforementioned third preset probability threshold can be a value preset according to actual needs.
[0042] The aforementioned third pre-set reliability threshold can be a value pre-set according to actual needs.
[0043] The aforementioned second warning level can be either a false alarm or a warning level.
[0044] In step S501, the above-mentioned determination of the warning level based on the third preset probability threshold, entry probability, third preset confidence threshold, and intent confidence can be as follows: if the entry probability is less than or equal to the third preset probability threshold, or if the intent confidence is less than or equal to the third preset confidence threshold, the false alarm level is taken as the second warning level; if the entry probability is greater than the third preset probability threshold and the intent confidence is greater than the third preset confidence threshold, the warning level is taken as the second warning level.
[0045] Step S502: Based on the second warning level and the preset level response strategy, determine the anti-pinch control scheme for the car window.
[0046] In step S502, the above-mentioned vehicle window anti-pinch control scheme is determined based on the second warning level and the preset level response strategy. If the second warning level is a false alarm, the first scheme of the preset level response strategy is used as the vehicle window anti-pinch control scheme. If the second warning level is a warning level, the second scheme of the preset level response strategy is used as the vehicle window anti-pinch control scheme.
[0047] This application sets lower probability and confidence thresholds specifically for high-speed driving scenarios, which can adapt to the recognition disturbances caused by image shaking during driving, relax the risk judgment conditions, capture brief limb insertion behaviors in a timely manner, prioritize driving safety, take into account the sensitivity of recognition under high-speed conditions, effectively prevent pinching accidents when the window is raised at high speed, and improve the reliability of window raising and lowering control.
[0048] In some embodiments, step S104 may include steps S601 to S602: Step S601: When the light intensity value is greater than or equal to a preset light intensity threshold and the vehicle speed is less than a preset vehicle speed threshold, a third warning level is determined based on a fourth preset probability threshold, a fifth preset probability threshold, an entry probability, a fourth preset confidence threshold, a fifth preset confidence value, and an intent confidence level. The fourth preset probability threshold is less than the fifth preset probability threshold, the fourth preset confidence threshold is less than the fifth preset confidence value, the fourth preset probability threshold is greater than the first preset probability threshold, the fourth preset confidence threshold is greater than the first preset confidence value, the fifth preset probability threshold is greater than the second preset probability threshold, the fifth preset confidence threshold is greater than the second preset confidence value, the fifth preset probability threshold is less than one, and the fifth preset confidence value is less than one. The aforementioned fourth preset probability threshold can be a value preset according to actual needs.
[0049] The aforementioned fifth preset probability threshold can be a value preset according to actual needs.
[0050] The aforementioned fourth preset reliability threshold can be a value preset according to actual needs.
[0051] The aforementioned fifth preset reliability threshold can be a value preset according to actual needs.
[0052] The aforementioned third warning level can be a false alarm level, a warning level, an intervention level, or a trap level.
[0053] In step S601, the calculation process for determining the third warning level based on the fourth preset probability threshold, the fifth preset probability threshold, the entry probability, the fourth preset confidence threshold, the fifth preset confidence value, and the intent confidence is similar to the calculation process for determining the first warning level based on the first preset probability threshold, the second preset probability threshold, the entry probability, the first preset confidence threshold, the second preset confidence value, and the intent confidence in step S301, and will not be repeated here.
[0054] Step S602: Based on the third warning level and the preset level response strategy, determine the anti-pinch control scheme for the car window.
[0055] In step S602, the calculation process for determining the anti-pinch control scheme for the car window based on the third warning level and the preset level response strategy is similar to the calculation process for determining the anti-pinch control scheme for the car window based on the first warning level and the preset level response strategy in step S302, and will not be repeated here.
[0056] This application targets well-lit, low-speed driving conditions and uses a two-level probability threshold and confidence threshold with higher overall values to classify warning levels. It fully utilizes the advantage of higher image recognition accuracy under good lighting conditions, raises the judgment threshold, effectively eliminates false intrusion behavior caused by debris and light shadows, reduces unnecessary anti-pinch intervention, and while ensuring the window's ability to prevent injury, reduces the probability of false triggering and improves the reliability of window lifting control.
[0057] In some embodiments, step S102 may include steps S701 to S704: Step S701: When the light intensity value is less than the preset light intensity threshold, acquire all first in-vehicle images within the first preset time period; The aforementioned first preset time period can be a value preset according to actual needs.
[0058] In step S701, the above-mentioned acquisition of all first in-vehicle images within the first preset time period can be achieved by using an installed in-vehicle camera to acquire in-vehicle images at a acquisition frequency preset according to actual needs, storing the acquired images in a storage area preset according to actual needs, and extracting the in-vehicle images within the first preset time period from the storage area as all first in-vehicle images within the first preset time period.
[0059] Step S702: Collect all the first in-vehicle images as an in-vehicle image sequence; Step S703: When the light intensity value is greater than or equal to the preset light intensity threshold, acquire all second in-vehicle images within the second preset time period, wherein the second preset time period is less than the second preset time period. The aforementioned second preset time period can be a value preset according to actual needs.
[0060] In step S703, the calculation process for obtaining all second in-vehicle images within the second preset time period is similar to the calculation process for obtaining all first in-vehicle images within the first preset time period in step S701, and will not be repeated here.
[0061] Step S704: Collect all the second in-vehicle images as an in-vehicle image sequence.
[0062] This application dynamically adjusts the image acquisition time period based on the light intensity. In low-light conditions, a longer acquisition period is used to construct the image sequence to ensure sufficient frame data to suppress noise interference. In strong light conditions, the acquisition period is shortened to reduce redundant images. This not only improves the stability of motion behavior recognition under low light conditions, but also improves the computing efficiency in good lighting conditions, balancing image recognition accuracy and program running speed. This provides image data adapted to different lighting environments for subsequent determination of limb intrusion behavior, thereby improving the reliability and efficiency of window lifting control.
[0063] Additionally, refer to Figure 2 One embodiment of this application provides an anti-pinch control system for automobile windows, including a data acquisition module 1100, an in-vehicle image sequence determination module 1200, an entry probability and intent confidence determination module 1300, and an anti-pinch control scheme determination module 1400, wherein: The data acquisition module 1100 is used to acquire environmental parameter values when it receives information that any window is raised. The in-vehicle image sequence determination module 1200 is used to determine the in-vehicle image sequence based on environmental parameter values; The entry probability and intent confidence determination module 1300 is used to determine the entry probability and intent confidence based on the in-vehicle image sequence. The entry probability is used to characterize the probability value of a preset human body part extending into the preset dangerous area of the window, and the intent confidence is used to characterize the confidence of the preset human body part extending into the window. The vehicle window anti-pinch control scheme determination module 1400 is used to determine the vehicle window anti-pinch control scheme based on environmental parameter values, entry probability and intent confidence.
[0064] This system collects environmental parameters simultaneously upon detecting a window raising command, and then matches the corresponding in-vehicle image sequence with these parameters for identification. It calculates the entry probability and behavioral intent confidence of human body parts in dangerous areas, and then combines environmental parameters, entry probability, and intent confidence to jointly determine the window anti-pinch control scheme. This effectively avoids image recognition deviations caused by changes in lighting, avoids misidentification problems caused by single visual judgment, distinguishes between accidental occlusion and genuine active insertion behavior, significantly reduces the probability of false triggering and missed detection of window anti-pinch, and improves the accuracy and reliability of window raising and lowering control.
[0065] It should be noted that the system embodiments described above are based on the same inventive concept as the method embodiments described above. Therefore, the relevant content of the method embodiments described above is also applicable to the system embodiments described above, and will not be repeated here.
[0066] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with relevant regulations. The acquisition, storage, use and processing of data in the technical solution of this application all comply with the relevant provisions of national laws and regulations.
[0067] like Figure 3 One embodiment of this application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described anti-pinch control method for automobile windows. The electronic device includes: At least one memory; At least one processor; At least one program; The program is stored in memory, and the processor executes at least one program to implement a car window anti-pinch control method according to the above embodiments of the present disclosure.
[0068] Electronic devices can be any smart terminal, including mobile phones, tablets, personal digital assistants (PDAs), and in-vehicle computers.
[0069] The electronic devices according to embodiments of this application will now be described in detail.
[0070] The processor 1600 can be implemented using a general-purpose central processing unit (CPU), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this disclosure. The memory 1700 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 1700 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 1700 and is called and executed by the processor 1600 to implement a car window anti-pinch control method according to an embodiment of this disclosure.
[0071] The input / output interface 1800 is used to implement information input and output. The communication interface 1900 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 2000 transmits information between various components of the device (e.g., processor 1600, memory 1700, input / output interface 1800, and communication interface 1900); The processor 1600, memory 1700, input / output interface 1800 and communication interface 1900 are connected to each other within the device via bus 2000.
[0072] This disclosure also provides a storage medium, which is a computer-readable storage medium storing computer-executable instructions for causing a computer to execute the above-described anti-pinch control method for automobile windows.
[0073] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
[0074] The above is a detailed description of the preferred embodiments of this application. However, the embodiments of this application are not limited to the above-described implementation methods. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the embodiments of this application. All such equivalent modifications or substitutions are included within the scope defined by the claims of the embodiments of this application.
Claims
1. A method of pinch control for an automotive vehicle window, characterized by, The vehicle window anti-pinch control method includes: Upon receiving notification that any window in the vehicle is raised, obtain environmental parameter values; Based on the environmental parameter values, a sequence of in-vehicle images is determined; Based on the in-vehicle image sequence, the entry probability and intent confidence are determined, wherein the entry probability is used to characterize the probability value of a preset human body part extending into the preset dangerous area of the vehicle window, and the intent confidence is used to characterize the confidence level of the preset human body part extending into the vehicle. Based on the environmental parameter values, the entry probability, and the intent confidence level, a vehicle window anti-pinch control scheme is determined.
2. The method of claim 1, wherein, The determination of entry probability and intent confidence based on the in-vehicle image sequence includes: The motion trajectory features of the in-vehicle image sequence are extracted using a temporal convolutional network model. The motion trajectory features include motion direction vector, motion speed, motion acceleration, and distance from the boundary between the preset human body part and the preset dangerous area of the vehicle window. The motion trajectory features are input into an attention-based intention prediction model to obtain the entry probability and the intention confidence output by the attention-based intention prediction model.
3. The method of claim 2, wherein, The environmental parameter values include vehicle speed and light intensity values. The process of determining the anti-pinch control scheme for the vehicle window based on the environmental parameter values, the entry probability, and the intent confidence level includes: When the light intensity value is less than a preset light intensity threshold and the vehicle speed is less than a preset vehicle speed threshold, a first warning level is determined based on a first preset probability threshold, a second preset probability threshold, the entry probability, a first preset confidence threshold, a second preset confidence value, and the intent confidence, wherein the first preset probability threshold is less than the second preset probability threshold, and the first preset confidence threshold is less than the second preset confidence value. Based on the first warning level and the preset level response strategy, the anti-pinch control scheme for the car window is determined.
4. The method for preventing pinching of automobile windows according to claim 3, characterized in that, The first warning level is a false alarm level, a warning level, an intervention level, or an anti-pinch level. Determining the first warning level based on a first preset probability threshold, a second preset probability threshold, the entry probability, a first preset confidence threshold, a second preset confidence value, and the intent confidence level includes: If the probability of entry is less than or equal to the first preset probability threshold, or if the confidence level of intent is less than or equal to the first preset confidence threshold, the false alarm level will be used as the first warning level. When the probability of entry is greater than the first preset probability threshold and less than the second preset probability threshold, and the confidence level of intent is greater than the first preset confidence threshold, the warning level is used as the first warning level; when the probability of entry is greater than the first preset probability threshold, and the confidence level of intent is greater than the first preset confidence threshold and less than the second preset confidence value, the warning level is used as the first warning level. If the probability of entry is greater than or equal to the second preset probability threshold and less than one, and the confidence level of intent is greater than or equal to the second preset confidence value, the intervention level shall be used as the first warning level. When the probability of entry is equal to one, the anti-pinch level is taken as the first warning level.
5. The method for preventing pinching of automobile windows according to claim 3, characterized in that, The process of determining an anti-pinch control scheme for vehicle windows based on the environmental parameter values, the entry probability, and the intent confidence level includes: When the vehicle speed is greater than or equal to the preset vehicle speed threshold, a second warning level is determined based on a third preset probability threshold, the entry probability, a third preset confidence threshold, and the intent confidence, wherein the third preset probability threshold is less than the first preset probability threshold, and the third preset confidence threshold is less than the first preset confidence value. Based on the second warning level and the preset level response strategy, the anti-pinch control scheme for the car window is determined.
6. The method for preventing pinching of automobile windows according to claim 5, characterized in that, The process of determining an anti-pinch control scheme for vehicle windows based on the environmental parameter values, the entry probability, and the intent confidence level includes: When the light intensity value is greater than or equal to the preset light intensity threshold and the vehicle speed is less than the preset vehicle speed threshold, a third warning level is determined based on a fourth preset probability threshold, a fifth preset probability threshold, the entry probability, a fourth preset confidence threshold, a fifth preset confidence value, and the intent confidence level. The fourth preset probability threshold is less than the fifth preset probability threshold, the fourth preset confidence threshold is less than the fifth preset confidence value, the fourth preset probability threshold is greater than the first preset probability threshold, the fourth preset confidence threshold is greater than the first preset confidence value, the fifth preset probability threshold is greater than the second preset probability threshold, the fifth preset confidence threshold is greater than the second preset confidence value, the fifth preset probability threshold is less than one, and the fifth preset confidence value is less than one. Based on the third warning level and the preset level response strategy, the anti-pinch control scheme for the car window is determined.
7. The method for preventing pinching of automobile windows according to claim 6, characterized in that, Determining the in-vehicle image sequence based on the environmental parameter values includes: When the light intensity value is less than a preset light intensity threshold, acquire all first in-vehicle images within a first preset time period; All the first in-vehicle images are used as the in-vehicle image sequence; When the light intensity value is greater than or equal to the preset light intensity threshold, all second in-vehicle images within a second preset time period are acquired, wherein the second preset time period is less than the second preset time period. All the second in-vehicle images are used as the in-vehicle image sequence.
8. A car window anti-pinch control system, characterized in that, The vehicle window anti-pinch control system includes: The data acquisition module is used to acquire environmental parameter values when receiving information that any window is raised. The in-vehicle image sequence determination module is used to determine the in-vehicle image sequence based on the environmental parameter values; An entry probability and intent confidence determination module is used to determine the entry probability and intent confidence based on the in-vehicle image sequence, wherein the entry probability is used to characterize the probability value of a preset human body part extending into a preset dangerous area of the vehicle window, and the intent confidence is used to characterize the confidence of the preset human body part extending into the vehicle. The vehicle window anti-pinch control scheme determination module is used to determine the vehicle window anti-pinch control scheme based on the environmental parameter values, the entry probability, and the intent confidence level.
9. An electronic device, characterized in that, It includes at least one processor and a memory for communicatively connecting with the processor; the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to cause the at least one processor to perform a vehicle window anti-pinch control method as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions for causing a computer to perform a vehicle window anti-pinch control method as described in any one of claims 1 to 7.
