Camera event triggering method, apparatus, device, and medium

Abstract

The application relates to the technical field of camera power consumption control, in particular to a camera event triggering method and device, equipment and medium, the method comprising the following steps: when an image acquisition module is in a low-power state, emitting a first ultrasonic detection signal, determining a first double-channel echo feature of a current event; determining a plurality of grade matching degrees corresponding to the current event; determining a target detection parameter, performing supplementary detection, and obtaining a second double-channel echo feature corresponding to the current event; after the second double-channel echo feature is obtained, determining a target event processing grade corresponding to the current event, and controlling the image acquisition module to remain in the low-power state or switch from the low-power state to an image acquisition state according to the target event processing grade. The application can solve the technical problems that a low-power camera in the prior art directly triggers an image acquisition module according to non-imaging detection information, inaccurate event judgment, a large number of invalid wake-up times and high overall power consumption are prone to occur.

Description

Technical Field

[0001] This application relates to the field of camera power consumption control technology, and in particular to a camera event triggering method, apparatus, device and medium. Background Technology

[0002] With the widespread use of cameras in home security, outdoor surveillance, and access control, battery-powered cameras are becoming increasingly common. Battery-powered cameras typically need to operate for extended periods with limited battery capacity, while the image acquisition unit consumes relatively high power during image acquisition. To extend the battery life of these cameras, the image acquisition unit is usually kept in a low-power state under normal conditions, only switching to image acquisition mode after an event requiring processing is detected.

[0003] Existing camera event triggering methods involve detecting targets around the camera using sensors such as infrared, radar, or ultrasound while the image acquisition device is in a low-power state. When the detection signal from these sensors reaches a trigger threshold, the camera is directly woken up, allowing the image acquisition device to acquire images. Compared to continuous operation of the image acquisition device, this triggering control method can reduce the camera's power consumption to some extent. However, the detection information obtained by non-imaging detection devices during low-power detection is usually limited. For example, when using ultrasound for event detection, the ultrasonic echo is easily affected by factors such as target position, target motion state, target reflection characteristics, environmental noise, and multipath reflection. Relying solely on a single trigger threshold makes it difficult to reliably determine the processing requirements corresponding to the current event. When the processing requirements are unclear, directly waking up the image acquisition device can easily lead to unnecessary image acquisition events causing the device to start ineffectively, and can also result in a high number of invalid camera wake-ups and high overall camera power consumption.

[0004] Therefore, there is an urgent need to design a method, device, equipment, and medium for triggering camera events. Summary of the Invention

[0005] Therefore, it is necessary to provide a camera event triggering method, device, equipment, and medium to address the above-mentioned technical problems. This can solve the technical problems in the prior art where low-power cameras directly trigger the image acquisition module based on non-imaging detection information, which easily leads to inaccurate event judgment, numerous invalid wake-up attempts, and high overall power consumption.

[0006] The technical solution of this invention is as follows: A camera event triggering method, applied to a camera, the camera including an image acquisition module, an ultrasonic transmitter, a first ultrasonic receiver, and a second ultrasonic receiver, the method comprising: When the image acquisition module is in a low-power state, the ultrasonic transmitter is controlled to emit a first ultrasonic detection signal, and the first dual-channel echo feature of the current event is determined based on the echo signals received by the first ultrasonic receiver and the second ultrasonic receiver. Based on a pre-established event echo feature library, the first dual-channel echo features are matched for level to determine multiple level matching degrees corresponding to the current event, wherein each level matching degree corresponds to an event processing level. When multiple matching degrees of the stated level do not meet the preset level determination conditions, the target detection parameters are determined, and supplementary detection is performed based on the target detection parameters to obtain the second dual-channel echo feature corresponding to the current event. After obtaining the second dual-channel echo feature, the target event processing level corresponding to the current event is determined based on the first dual-channel echo feature and the second dual-channel echo feature, and the image acquisition module is controlled to maintain a low-power state or switch from a low-power state to an image acquisition state based on the target event processing level.

[0007] Optionally, a camera event triggering device is also provided, applied to a camera, the camera including an image acquisition module, an ultrasonic transmitter, a first ultrasonic receiver, and a second ultrasonic receiver, the device comprising: The echo feature determination module is used to control the ultrasonic transmitter to emit a first ultrasonic detection signal when the image acquisition module is in a low power state, and to determine the first dual-channel echo feature of the current event based on the echo signals received by the first ultrasonic receiver and the second ultrasonic receiver. The level matching determination module is used to perform level matching on the first dual-channel echo features based on a pre-established event echo feature library, and determine multiple level matching degrees corresponding to the current event, wherein each level matching degree corresponds to an event processing level. The supplementary detection execution module is used to determine the target detection parameters when multiple level matching degrees do not meet the preset level determination conditions, and to perform supplementary detection based on the target detection parameters to obtain the second dual-channel echo feature corresponding to the current event. The power consumption state control module is used to determine the target event processing level corresponding to the current event based on the first dual-channel echo feature and the second dual-channel echo feature after obtaining the second dual-channel echo feature, and control the image acquisition module to maintain a low power consumption state or switch from a low power consumption state to an image acquisition state based on the target event processing level.

[0008] Optionally, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps described in the camera event triggering method above.

[0009] Optionally, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed by a processor, implements the steps described in the camera event triggering method above.

[0010] The technical effects achieved by this invention are as follows: The aforementioned camera event triggering method, apparatus, device, and medium, when the image acquisition module is in a low-power state, emits a first ultrasonic detection signal via an ultrasonic transmitter and receives the corresponding echo signal using a first and second ultrasonic receiver to determine the first dual-channel echo feature corresponding to the current event. This allows for the acquisition of multi-channel ultrasonic echo information of the current event without activating the image acquisition module, reducing the power consumption of the image acquisition module due to continuous event detection. The first dual-channel echo feature is then matched using a pre-established event echo feature library to determine multiple matching degrees corresponding to the current event, with each matching degree corresponding to an event processing level. This converts the ultrasonic echo information of the current event into quantitative matching results corresponding to different event processing levels. This not only determines the possible event processing level of the current event but also assesses whether the initial detection result is sufficient to determine the processing level based on the relationship between multiple matching degrees. When multiple matching degrees do not meet the preset level determination conditions, target detection parameters are determined, and supplementary detection is performed according to these parameters to obtain the second dual-channel echo feature corresponding to the current event. This system implements supplementary detection only when the initial level matching result is unclear, avoiding multiple ultrasonic detections for all events. Simultaneously, supplementary echo information is obtained using target detection parameters, increasing the detection criteria for distinguishing the current event processing level and reducing the possibility of misjudgment due to environmental noise, target position, and multipath reflections affecting single detection results. After obtaining the second dual-channel echo characteristics, the target event processing level corresponding to the current event is determined based on the first and second dual-channel echo characteristics. The echo information obtained from both detections is used to jointly determine the processing requirements of the current event, reducing the impact of accidental changes in a single echo signal on the event determination result and improving the reliability of the target event processing level determination. Finally, the image acquisition module is controlled to maintain a low-power state or switch from a low-power state to an image acquisition state based on the target event processing level, implementing differentiated camera state control for events with different processing requirements. For events that do not require image acquisition, the image acquisition module remains in a low-power state; for events that require image acquisition, the image acquisition module is switched to the image acquisition state. This reduces invalid wake-ups of the image acquisition module caused by non-target events or events with unclear judgments, lowers the power consumption of the image acquisition module, and extends the battery life of the battery-powered camera. Furthermore, this application can improve the reliability of camera event triggering by controlling the camera state while maintaining low power consumption in the image acquisition module, reducing invalid wake-ups and invalid image acquisitions, thus balancing event detection reliability and overall camera power consumption. Attached Figure Description

[0011] Figure 1This is a flowchart illustrating a camera event triggering method in one embodiment; Figure 2 This is a structural block diagram of a camera event triggering device in one embodiment. Detailed Implementation

[0012] In the following description, specific details such as particular system architectures and technologies are set forth for illustrative purposes and not for limiting purposes, in order to provide a thorough understanding of the embodiments of this application.

[0013] In one embodiment, such as Figure 1 As shown, a camera event triggering method is provided, applied to a camera, the camera including an image acquisition module, an ultrasonic transmitter, a first ultrasonic receiver, and a second ultrasonic receiver, the method including: Step S100: When the image acquisition module is in a low power state, control the ultrasonic transmitter to emit a first ultrasonic detection signal, and determine the first dual-channel echo feature of the current event based on the echo signals received by the first ultrasonic receiver and the second ultrasonic receiver. Step S200: Based on the pre-established event echo feature library, perform level matching on the first dual-channel echo features to determine multiple level matching degrees corresponding to the current event, wherein each level matching degree corresponds to an event processing level; Step S300: When multiple level matching degrees do not meet the preset level determination conditions, determine the target detection parameters, and obtain the second dual-channel echo feature corresponding to the current event based on the target detection parameters; In this step, specifically, it refers to controlling the ultrasonic transmitter to work based on the target detection parameters, and controlling the first ultrasonic receiver and the second ultrasonic receiver to receive signals, thereby realizing the execution of supplementary detection, and obtaining the second dual-channel echo characteristics corresponding to the current event through the first ultrasonic receiver and the second ultrasonic receiver; Step S400: After obtaining the second dual-channel echo feature, determine the target event processing level corresponding to the current event based on the first dual-channel echo feature and the second dual-channel echo feature, and control the image acquisition module to maintain a low-power state or switch from a low-power state to an image acquisition state based on the target event processing level.

[0014] In this embodiment, when the image acquisition module is in a low-power state, a first ultrasonic detection signal is emitted through an ultrasonic transmitter, and the corresponding echo signal is received using a first ultrasonic receiver and a second ultrasonic receiver to determine the first dual-channel echo feature corresponding to the current event. This allows for the acquisition of multi-channel ultrasonic echo information of the current event without activating the image acquisition module, reducing the power consumption of the image acquisition module due to continuous event detection. The first dual-channel echo feature is then matched using a pre-established event echo feature library to determine multiple matching degrees corresponding to the current event, with each matching degree corresponding to an event processing level. This converts the ultrasonic echo information of the current event into quantized matching results corresponding to different event processing levels. This not only determines the possible event processing level of the current event but also assesses whether the initial detection result is sufficient to determine the processing level based on the relationship between multiple matching degrees. When multiple matching degrees do not meet the preset level determination conditions, target detection parameters are determined, and supplementary detection is performed according to these parameters to obtain the second dual-channel echo feature corresponding to the current event. This system implements supplementary detection only when the initial level matching result is unclear, avoiding multiple ultrasonic detections for all events. Simultaneously, supplementary echo information is obtained using target detection parameters, increasing the detection criteria for distinguishing the current event processing level and reducing the possibility of misjudgment due to environmental noise, target position, and multipath reflections affecting single detection results. After obtaining the second dual-channel echo characteristics, the target event processing level corresponding to the current event is determined based on the first and second dual-channel echo characteristics. The echo information obtained from both detections is used to jointly determine the processing requirements of the current event, reducing the impact of accidental changes in a single echo signal on the event determination result and improving the reliability of the target event processing level determination. Finally, the image acquisition module is controlled to maintain a low-power state or switch from a low-power state to an image acquisition state based on the target event processing level, implementing differentiated camera state control for events with different processing requirements. For events that do not require image acquisition, the image acquisition module remains in a low-power state; for events that require image acquisition, the image acquisition module is switched to the image acquisition state. This reduces invalid wake-ups of the image acquisition module caused by non-target events or events with unclear judgments, lowers the power consumption of the image acquisition module, and extends the battery life of the battery-powered camera. Furthermore, this application can improve the reliability of camera event triggering by controlling the camera state while maintaining low power consumption in the image acquisition module, reducing invalid wake-ups and invalid image acquisitions, thus balancing event detection reliability and overall camera power consumption.

[0015] In one embodiment, step S100 involves controlling the ultrasonic transmitter to emit a first ultrasonic detection signal, and determining a first dual-channel echo characteristic of the current event based on the echo signals received by the first and second ultrasonic receivers, including: Step S110: Control the ultrasonic transmitter to emit a first ultrasonic detection signal according to the first detection parameters; Step S120: Obtain the first echo signal corresponding to the current event through the first ultrasonic receiver; Step S130: Obtain the second echo signal corresponding to the current event through the second ultrasonic receiver; Step S140: Determine the first dual-channel echo feature corresponding to the current event based on at least two channel association features between the first echo signal and the second echo signal.

[0016] In this embodiment, the camera further includes a controller, which is communicatively connected to the image acquisition module, the ultrasonic transmitter, the first ultrasonic receiver, and the second ultrasonic receiver. The first and second ultrasonic receivers are spaced apart and form independent echo receiving channels. The controller controls the transmission of ultrasonic detection signals, synchronously samples the two echo receiving channels, and determines the first dual-channel echo characteristic corresponding to the current event based on the echo signals obtained from the two echo receiving channels.

[0017] The image acquisition module being in a low-power state means that the image sensor, image signal processor, or at least some of the circuitry related to image acquisition is in a state of sleep, power-off, low-frequency operation, or cessation of image output. While the image acquisition module is in a low-power state, at least some of the low-power processing circuitry in the ultrasonic transmitter, first ultrasonic receiver, second ultrasonic receiver, and controller remains operational to perform initial detection of the current event.

[0018] In step S110, the first detection parameters include at least one of the following: transmission frequency, number of transmission pulses, duration of a single pulse, pulse repetition interval, signal encoding method, and transmission power.

[0019] The first detection parameter is the detection parameter preset during the camera's factory calibration process. During the camera's factory calibration phase, ultrasonic detection is performed on calibration targets at different preset detection positions according to multiple sets of candidate detection parameters. The echo signals obtained under each set of candidate detection parameters and the corresponding detection energy consumption are recorded. The set of candidate detection parameters that can obtain effective echo signals within the camera's preset detection range and whose detection energy consumption meets the camera's low power consumption requirements is determined as the first detection parameter.

[0020] The controller sends a first transmission control command to the ultrasonic transmitter, causing the ultrasonic transmitter to continuously transmit a preset number of ultrasonic pulses at a preset transmission frequency to form a first ultrasonic detection signal.

[0021] When the controller controls the ultrasonic transmitter to emit the first ultrasonic detection signal, it synchronously records the start time of the emission and uses the start time of the emission as a common time reference for the first and second ultrasonic receivers to receive the echo.

[0022] In step S120, after the first ultrasonic detection signal propagates to the target around the camera, it is reflected by the target to form an ultrasonic echo. The first ultrasonic receiver receives the ultrasonic echo and obtains the first original echo signal corresponding to the first receiving channel.

[0023] The controller opens a first receiving time window at a preset time after the first ultrasonic detection signal is transmitted, and samples the first raw echo signal within the first receiving time window. The controller performs signal processing on the first raw echo signal to obtain the first echo signal corresponding to the current event.

[0024] In step S130, the second ultrasonic receiver receives the ultrasonic echo formed by the reflection of the same target and obtains the second original echo signal corresponding to the second receiving channel.

[0025] The controller samples the second original echo signal using the same sampling frequency and receiving time window as the first original echo signal, and performs signal processing to obtain the second echo signal corresponding to the current event.

[0026] In step S140, the channel correlation feature is used to characterize the relative relationship between the first echo signal and the second echo signal.

[0027] Channel correlation characteristics include at least two of the following: echo arrival time difference, echo amplitude ratio, and echo energy ratio.

[0028] Specifically, the controller extracts one or more echo peaks from the first echo signal and the second echo signal, respectively. When the first echo signal includes multiple echo peaks, the controller determines the echo peak with the largest amplitude as the first target echo peak; when the first echo signal includes only one echo peak, the controller determines that echo peak as the first target echo peak.

[0029] Similarly, the controller determines the second target echo peak with the largest amplitude from the second echo signal, and pairs the first target echo peak with the second target echo peak to form a target echo peak pair.

[0030] The controller compares the occurrence times of the two target echo peaks in the target echo peak pair and determines the difference between their occurrence times as the echo arrival time difference. The echo arrival time difference is used to characterize the difference in arrival times of the main echo components in the first echo receiving channel and the second echo receiving channel.

[0031] The controller also compares the amplitudes of the first target echo peak and the second target echo peak, determining the relative amplitude relationship between the two target echo peaks as the echo amplitude ratio. The echo amplitude ratio reflects the difference in reflection intensity of the target relative to the two echo receiving channels.

[0032] In other embodiments, the controller determines the echo segment including the first target echo peak as the first target echo segment and the echo segment including the second target echo peak as the second target echo segment, and determines the echo energy of the first target echo segment and the second target echo segment respectively, and determines the echo energy ratio based on the relative relationship between the two echo energies.

[0033] In one implementation, the controller selects the echo arrival time difference and the echo amplitude ratio as two channel association features, and combines the two channel association features in a preset order to form a first dual-channel echo feature.

[0034] Of course, in other embodiments, those skilled in the art can also configure the controller to select echo arrival time difference, echo amplitude ratio and waveform correlation as three channel association features, and combine the three channel association features to form a first dual-channel echo feature.

[0035] For example, the target echo peak in the first echo signal appears earlier than the target echo peak in the second echo signal, with an arrival time difference of 0.22 ms between the two target echo peaks; the amplitude of the target echo peak in the first echo signal is 1.2 times the amplitude of the target echo peak in the second echo signal; and the waveform correlation between the first and second target echo segments is 0.84. The controller combines the above arrival time difference, amplitude relative relationship, and waveform correlation to obtain the first dual-channel echo characteristic corresponding to the current event.

[0036] To reduce the impact of differences in the numerical range and dimensions of different channel association features on subsequent level matching, the controller also normalizes channel association features such as echo arrival time difference, echo amplitude ratio, and echo energy ratio, and arranges at least two normalized channel association features in a preset order to form the first dual-channel echo feature.

[0037] This embodiment directly utilizes the correlation between the first and second echo signals to construct the first dual-channel echo feature, without needing to first calculate the target's two-dimensional coordinates based on the ultrasonic wave propagation time. This reduces the impact of target coordinate calculation errors on subsequent event level matching and preserves echo difference information between the two echo receiving channels that can be used to distinguish different events.

[0038] Furthermore, while the image acquisition module maintains a low-power state, the first dual-channel echo feature corresponding to the current event is obtained using two ultrasonic receiving channels, providing a data foundation for subsequently determining the matching degree of multiple levels corresponding to the current event based on the event echo feature library.

[0039] In one embodiment, the event echo feature library is pre-established, including: Step S201: Obtain the historical detection parameters, historical first echo signal, historical second echo signal, and historical event images corresponding to multiple historical events; Step S202: Determine the historical dual-channel echo characteristics corresponding to the historical event based on at least two channel association characteristics between the historical first echo signal and the historical second echo signal corresponding to the same historical event; Step S203: Perform target recognition on the historical event image to obtain the historical image recognition result corresponding to the historical event; Step S204: Based on the historical image recognition results and the preset event processing strategy, determine the event processing level corresponding to the historical event, wherein different event processing levels correspond to different camera control methods; Step S205: Associate and store the historical detection parameters, historical dual-channel echo characteristics, and event processing level corresponding to the same historical event to establish an event echo feature library.

[0040] In this embodiment, the event echo feature library is pre-established before performing level matching on the current event. The event echo feature library is established during the camera's factory calibration phase or historical operation phase. The camera also includes a memory, in which the controller stores the acquired historical detection parameters, historical dual-channel echo features, and event processing levels to form the event echo feature library. The memory is a local memory located within the camera.

[0041] In step S201, historical events refer to events recorded jointly by ultrasonic detection and image acquisition during the establishment of the event echo feature library. Historical events are events simulated during the camera's factory calibration phase. Alternatively, historical events are events actually acquired during the camera's installation, debugging, or historical operation phases.

[0042] Historical detection parameters refer to the detection parameters actually used by the ultrasonic transmitter, the first ultrasonic receiver, and the second ultrasonic receiver when acquiring the corresponding historical first echo signal and historical second echo signal. Historical detection parameters are set according to the actual needs of those skilled in the art and include at least one of the following: transmission frequency, number of transmitted pulses, and duration of a single pulse.

[0043] Before establishing the event echo feature library, multiple candidate detection parameters are pre-set. At least one of these candidate detection parameters must differ. The controller controls the ultrasonic transmitter, the first ultrasonic receiver, and the second ultrasonic receiver to perform historical detection according to these candidate detection parameters. The candidate detection parameters actually used for historical detection are recorded as the historical detection parameters for the corresponding historical events.

[0044] For the same historical event, the controller sequentially performs ultrasonic detection according to multiple candidate detection parameters. For each candidate detection parameter, the controller controls the ultrasonic transmitter to emit the corresponding historical ultrasonic detection signal, and receives the ultrasonic echo formed by the reflection from the same target through the first ultrasonic receiver and the second ultrasonic receiver, respectively, to obtain the historical first original echo signal and the historical second original echo signal corresponding to the candidate detection parameter.

[0045] The controller performs signal processing on the first and second historical raw echo signals respectively to obtain the first and second historical echo signals corresponding to the historical events. The signal processing includes at least one of signal amplification, bandpass filtering, analog-to-digital conversion, baseline correction, and noise suppression.

[0046] When performing historical ultrasonic detection, the controller also keeps the image acquisition module in image acquisition mode to obtain historical event images of the corresponding historical events. For the same historical event, while the image acquisition module is in image acquisition mode, ultrasonic detection is performed sequentially according to multiple candidate detection parameters. Therefore, it is not necessary to start the image acquisition module separately for each candidate detection parameter, but to use the recognition results of the same historical event image to mark the historical dual-channel echo features obtained for the historical event under different historical detection parameters.

[0047] The controller assigns a historical event identifier to each historical event and associates the acquisition times of historical detection parameters, historical first echo signals, historical second echo signals, and historical event images with the same historical event identifier. The controller also limits the time difference between the ultrasonic detection time and the image acquisition time within a preset time range, ensuring that the historical first echo signals, historical second echo signals, and historical event images correspond to the same target or the same scene change. This preset time range is determined during the camera's factory calibration phase, based on the time required to complete historical detection using multiple candidate detection parameters sequentially and the movement speed of the calibrated target.

[0048] In step S202, the controller extracts one or more historical echo peaks from the historical first echo signal and the historical second echo signal, respectively. When the historical first echo signal includes multiple historical echo peaks, the controller determines the historical echo peak with the largest amplitude as the historical first target echo peak; when the historical first echo signal includes only one historical echo peak, the controller determines that historical echo peak as the historical first target echo peak.

[0049] Similarly, the controller determines the historical second target echo peak with the largest amplitude from the historical second echo signal, and pairs the historical first target echo peak with the historical second target echo peak to form a target historical echo peak pair. The selection method for historical echo peaks is the same as the selection method for echo peaks corresponding to the current event, so that the historical dual-channel echo characteristics and the first dual-channel echo characteristics have consistent formation rules.

[0050] The controller compares the occurrence times of two historical target echo peaks in the target historical echo peak pair to obtain the historical echo arrival time difference; compares the amplitudes of the two historical target echo peaks to obtain the historical echo amplitude ratio; or, determines the echo energy of the historical echo segment where the two historical target echo peaks are located respectively, and obtains the historical echo energy ratio based on the relative relationship between the two echo energies.

[0051] When neither the first historical echo signal nor the second historical echo signal has a historical echo peak value extracted, the controller marks the detection result of the corresponding historical event under the current historical detection parameters as having no valid echo and stops storing the historical dual-channel echo feature corresponding to the detection result into the event echo feature library.

[0052] The channel correlation feature type used in the historical dual-channel echo features is the same as that used in the first dual-channel echo features corresponding to the current event. For example, when the first dual-channel echo features include echo arrival time difference and echo amplitude ratio, the historical dual-channel echo features also include historical echo arrival time difference and historical echo amplitude ratio. This allows the first dual-channel echo features corresponding to the current event to be compared with the historical dual-channel echo features.

[0053] In addition, the controller also normalizes the channel correlation features such as historical echo arrival time difference, historical echo amplitude ratio and historical echo energy ratio according to the same normalization rule as the first dual-channel echo feature, and arranges at least two channel correlation features after normalization in a preset order to form historical dual-channel echo features.

[0054] When ultrasonic detection is performed on the same historical event using multiple historical detection parameters, the controller determines the corresponding historical dual-channel echo characteristics based on the historical first echo signal and historical second echo signal corresponding to each historical detection parameter. Therefore, the same historical event corresponds to multiple historical dual-channel echo characteristics, and each historical dual-channel echo characteristic corresponds to a historical detection parameter actually used.

[0055] In step S203, the historical event images include multiple frames of images continuously acquired within a preset image acquisition time period. The controller determines at least one of the target's movement direction, movement distance, and dwell time based on the positional changes of the target area in the multiple frames of images.

[0056] Specifically, the controller acquires a background reference image captured by the camera when no moving target is present, and performs differential processing on the historical event image and the background reference image to obtain the changed regions in the historical event image. The controller also performs inter-frame differential processing on adjacent image frames in the historical event image, and determines the motion region based on the pixel changes between adjacent image frames.

[0057] The controller performs denoising and binarization on the changed or moving regions to determine candidate target regions in the historical event images. The controller deletes candidate target regions with an area smaller than a preset area threshold to reduce the impact of image noise, lighting changes, flying insects, or short-term local disturbances on the historical image recognition results.

[0058] The controller extracts at least one image feature from the candidate target region, including region area, contour height, and contour aspect ratio. Multiple feature determination conditions are pre-set for each target category. These feature determination conditions include at least one of the following: target region area range, contour height range, contour width range, and contour aspect ratio range.

[0059] It should be noted that the preset area threshold, the feature determination conditions corresponding to multiple target categories, the preset valid target conditions, and the preset valid recognition conditions are all determined in advance by those skilled in the art based on calibration images with known target categories during the camera's factory calibration stage. The calibration images include multiple frames corresponding to people, pets, vehicles, no valid targets, and environmental disturbances, covering situations where the target is in different positions and at different distances. This application does not impose specific limitations.

[0060] The controller compares the image features of candidate target regions with feature determination conditions corresponding to multiple target categories, and determines the target category corresponding to the candidate target region as the target category whose image features meet the corresponding feature determination conditions. For example, when the contour height and contour aspect ratio of the candidate target region match the pre-stored standard personnel conditions, the feature determination conditions are met, and the controller identifies the candidate target region as a person; similarly, it can also identify the candidate target region as a pet or a vehicle based on different conditions.

[0061] When a candidate target region does not meet the feature determination conditions corresponding to a preset target category such as a person, pet, or vehicle, but meets the preset valid target conditions, the controller identifies the candidate target region as another target. When there is no candidate target region in the historical event image that meets the preset valid target conditions, the controller determines the historical image recognition result as no valid target.

[0062] Therefore, the results of historical image recognition include at least one of the following: target category, number of targets, region where the target is located, direction of target movement, and whether the target has entered the region of interest.

[0063] In other implementations, during the camera installation or debugging phase, the operator sets the area range, aspect ratio range, movement speed range, and area of ​​interest corresponding to the target category through the camera's interactive interface. The controller identifies historical event images based on the feature judgment conditions set by the operator, without the need for a pre-trained image recognition model.

[0064] In another embodiment, when a candidate target region in a historical event image is occluded, or when the image features extracted from the historical event image do not meet the preset recognition validity conditions, the controller determines the historical event as an invalid historical event and stops storing the corresponding historical dual-channel echo features into the event echo feature library. This reduces the impact of historical events with unclear target recognition results on the level labeling results in the event echo feature library.

[0065] In step S204, the preset event processing strategy is used to define the correspondence between different historical image recognition results and different event processing levels, as well as the correspondence between different event processing levels and different camera control methods. The preset event processing strategy is pre-configured when the camera leaves the factory.

[0066] Specifically, the preset event processing strategy includes at least one of target category conditions, target location conditions, and target movement direction conditions. The controller compares the historical image recognition results with the conditions in the preset event processing strategy to determine the event processing conditions satisfied by the historical event, and determines the event processing level corresponding to the event processing condition as the event processing level corresponding to the historical event. For example, when the historical image recognition result indicates that a person has entered the camera's area of ​​interest, and the preset event processing strategy stipulates that image acquisition is required when a person enters the area of ​​interest, the controller determines the historical event as the first event processing level. The camera control method corresponding to the first event processing level is to control the image acquisition module to switch from a low-power state to an image acquisition state and acquire the corresponding event image. When the historical image recognition result indicates that the target has not entered the area of ​​interest, or the target belongs to a preset non-interested target, and the preset event processing strategy stipulates that image acquisition is not required for this type of event, the controller determines the historical event as the second event processing level. The camera control method corresponding to the second event processing level is to control the image acquisition module to maintain a low-power state.

[0067] In other embodiments, three or more event processing levels are provided. For example, the first event processing level corresponds to controlling the image acquisition module to perform high-resolution or long-term image acquisition; the second event processing level corresponds to controlling the image acquisition module to perform low-frame-rate or short-term image acquisition; and the third event processing level corresponds to controlling the image acquisition module to maintain a low-power state.

[0068] In step S205, the controller establishes a historical echo feature record for each historical dual-channel echo feature. Each historical echo feature record includes at least a historical event identifier, historical detection parameters, historical dual-channel echo features, and event processing level. The historical echo feature record also includes at least one of the following: historical image recognition result, historical event image acquisition time, ultrasonic detection time, and historical event validity marker.

[0069] For the same historical event, multiple historical dual-channel echo features are obtained under different historical detection parameters. The controller establishes multiple historical echo feature records. The multiple historical echo feature records have the same historical event identifier and event processing level, but they have different historical detection parameters and corresponding historical dual-channel echo features.

[0070] As the number of historical events increases, the event echo feature library contains multiple historical dual-channel echo features corresponding to different historical detection parameters and different event processing levels. The controller classifies and stores these multiple historical dual-channel echo features according to historical detection parameters and event processing levels, so that it can subsequently select historical dual-channel echo features from the event echo feature library that can be compared with the first dual-channel echo feature based on the first detection parameter corresponding to the current event.

[0071] Historical event images are primarily used to determine the historical image recognition results corresponding to historical events. After determining the historical image recognition results and event processing level, the historical event images are saved for subsequent verification or re-determination of the event processing level; alternatively, only the historical image recognition results and event processing level can be saved, while the historical event images are deleted to reduce the storage space occupied by the event echo feature library.

[0072] This embodiment stores historical detection parameters, historical dual-channel echo characteristics, and event processing levels in a correlated manner, thereby providing a historical data foundation for subsequent level matching of current events and determination of target detection parameters.

[0073] In one embodiment, step S200: performing level matching on the first dual-channel echo features based on a pre-established event echo feature library to determine multiple level matching degrees corresponding to the current event, including: Step S210: Based on the first detection parameter corresponding to the first ultrasonic detection signal, determine the historical dual-channel echo features in the event echo feature library that satisfy the preset parameter correspondence condition between the historical detection parameters associated with the first detection parameter and the first detection parameter as candidate historical dual-channel echo features; Step S220: Compare the first dual-channel echo feature with each of the candidate historical dual-channel echo features to determine the feature matching degree between the first dual-channel echo feature and each of the candidate historical dual-channel echo features; Step S230: According to the event processing level associated with each candidate historical dual-channel echo feature, classify the multiple feature matching degrees to obtain multiple feature matching degree sets, wherein each feature matching degree set corresponds to an event processing level; Step S240: Perform statistical processing on the feature matching degrees in the multiple feature matching degree sets according to preset statistical rules, and determine the level matching degree of the current event corresponding to each of the event processing levels.

[0074] In this embodiment, in step S210, the first detection parameter refers to the detection parameter used by the controller when controlling the ultrasonic transmitter to emit the first ultrasonic detection signal and controlling the first ultrasonic receiver and the second ultrasonic receiver to receive the corresponding echo signal.

[0075] The controller filters multiple historical echo feature records in the event echo feature database based on the first detection parameter. Specifically, the controller compares each parameter item in the first detection parameter with the historical detection parameters in the historical echo feature records and determines whether the first detection parameter and the historical detection parameter meet the preset parameter correspondence conditions.

[0076] The preset parameter correspondence condition is used to determine whether the first detection parameter and the historical detection parameters are comparable. The preset parameter correspondence condition is that each parameter item in the first detection parameter and the historical detection parameters are the same, or that the difference between the corresponding parameter items in the first detection parameter and the historical detection parameters is within the preset parameter difference range. This preset parameter difference range is determined during the camera's factory calibration phase by performing ultrasonic detection on the same calibration target using multiple sets of detection parameters with similar parameters. For example, by comparing the dual-channel echo features obtained under different detection parameters, when the change in dual-channel echo features caused by the change in detection parameters does not affect the feature matching result of the same calibration target, the difference range between the corresponding detection parameters is determined as the preset parameter difference range.

[0077] In one implementation, multiple candidate detection parameters are each assigned a parameter identifier. When the controller performs an initial detection using one of the candidate detection parameters, it designates that candidate detection parameter as the first detection parameter and selects historical detection parameters with the same parameter identifier from the event echo feature library based on the parameter identifier of the first detection parameter. The controller then designates the historical dual-channel echo features associated with the selected historical detection parameters as candidate historical dual-channel echo features.

[0078] In another implementation, when there is no historical detection parameter in the event echo feature library that is exactly the same as the first detection parameter, the controller compares the transmission frequency, the number of transmission pulses, and the duration of a single pulse in the first detection parameter and each historical detection parameter, and determines the historical detection parameter whose difference between the corresponding parameter items is within the preset parameter difference range as the historical detection parameter that meets the preset parameter correspondence condition with the first detection parameter.

[0079] Since changes in historical detection parameters may cause changes in echo amplitude, echo energy, and echo duration, prioritizing the selection of historical dual-channel echo features corresponding to historical detection parameters that are the same as or similar to the first detection parameter can reduce the impact of differences in detection parameters on feature matching results.

[0080] In step S220, the first dual-channel echo feature and the candidate historical dual-channel echo feature use the same channel-related feature type and feature arrangement order. For example, when the first dual-channel echo feature includes echo arrival time difference and echo amplitude ratio, the candidate historical dual-channel echo feature also includes historical echo arrival time difference and historical echo amplitude ratio.

[0081] Specifically, the controller compares the channel-related features in the first dual-channel echo features with the corresponding channel-related features in the candidate historical dual-channel echo features according to the feature arrangement order, and determines the feature differences between the channel-related features.

[0082] The controller pre-stores the sub-matching degrees corresponding to different feature difference ranges. The smaller the feature difference, the higher the corresponding sub-matching degree; the larger the feature difference, the lower the corresponding sub-matching degree. The controller determines the sub-matching degree corresponding to each channel's associated feature based on the feature differences between the associated features of each channel.

[0083] The correspondence between different feature difference ranges and sub-item matching degrees is determined during the camera's factory calibration phase based on multiple calibration events with known event processing levels. Specifically, feature differences between dual-channel echo features belonging to the same event processing level and between dual-channel echo features belonging to different event processing levels are compared, and multiple feature difference ranges are set based on the distribution range of the two types of feature differences. Feature difference ranges appearing in the set of dual-channel echo features of the same event processing level correspond to higher sub-item matching degrees, while feature difference ranges appearing in the set of dual-channel echo features of different event processing levels correspond to lower sub-item matching degrees.

[0084] The controller performs statistical processing on multiple sub-item matching degrees according to preset feature weights to obtain the feature matching degree between the first dual-channel echo feature and a candidate historical dual-channel echo feature. The preset feature weights are pre-set based on the degree of influence of different channel association features on the event differentiation results. When no different preset feature weights are set, the controller averages the multiple sub-item matching degrees with the same weight.

[0085] Feature matching degree is used to represent the similarity between the first dual-channel echo feature and the corresponding candidate historical dual-channel echo feature. The higher the feature matching degree, the more similar the first dual-channel echo feature of the current event is to the corresponding candidate historical dual-channel echo feature; the lower the feature matching degree, the greater the difference between the two.

[0086] In step S230, the controller reads the historical echo feature records corresponding to each candidate historical dual-channel echo feature and obtains the corresponding event processing level from the historical echo feature records. The controller assigns the feature matching degrees corresponding to candidate historical dual-channel echo features associated with the same event processing level to the same feature matching degree set. For example, when the event echo feature library includes a first event processing level, a second event processing level, and a third event processing level, the controller assigns the feature matching degrees corresponding to candidate historical dual-channel echo features associated with the first event processing level to the first feature matching degree set, assigns the feature matching degrees corresponding to candidate historical dual-channel echo features associated with the second event processing level to the second feature matching degree set, and assigns the feature matching degrees corresponding to candidate historical dual-channel echo features associated with the third event processing level to the third feature matching degree set. Furthermore, each feature matching degree set corresponds to an event processing level, and each feature matching degree set includes the feature matching degree between the first dual-channel echo feature of the current event and one or more candidate historical dual-channel echo features under that event processing level.

[0087] In step S240, the preset statistical rule is to average multiple feature matching degrees from the same feature matching degree set. The controller determines the average result as the level matching degree of the current event corresponding to the event processing level associated with that feature matching degree set. For example, if the first feature matching degree set includes three feature matching degrees of 90%, 86%, and 82%, the controller averages these three degrees and determines the resulting 86% as the level matching degree of the current event corresponding to the first event processing level. If the second feature matching degree set includes three feature matching degrees of 78%, 75%, and 72%, the controller averages these three degrees as the level matching degree of the current event corresponding to the second event processing level. If the third feature matching degree set includes three feature matching degrees of 35%, 30%, and 25%, the controller averages these three degrees as the level matching degree of the current event corresponding to the third event processing level. This allows the controller to obtain multiple level matching degrees for the current event corresponding to the first, second, and third event processing levels, respectively. Multiple matching levels can represent the degree of similarity between the current event and historical events corresponding to different event handling levels.

[0088] This embodiment can improve the comparability between the first dual-channel echo characteristics and historical dual-channel echo characteristics.

[0089] In one embodiment, step S300: when multiple level matching degrees do not meet preset level determination conditions, determine target detection parameters, including: Step S310: When multiple level matching degrees do not meet the preset level determination conditions, at least two event processing levels that meet the preset candidate conditions are determined as event processing levels to be distinguished based on the multiple level matching degrees. Step S320: For each candidate detection parameter among multiple candidate detection parameters, obtain the historical dual-channel echo features corresponding to the candidate detection parameter and associated with each of the event processing levels to be distinguished from the event echo feature library, and obtain the historical echo feature set corresponding to each of the event processing levels to be distinguished. Step S330: Determine the representative feature values ​​of at least two channel association features in each of the historical echo feature sets, and determine the feature discrimination degree corresponding to each candidate detection parameter based on the difference between the representative feature values ​​corresponding to different event processing levels to be distinguished. Step S340: Determine the detection energy consumption corresponding to each candidate detection parameter, and determine the target detection parameter from the multiple candidate detection parameters based on the feature discrimination and detection energy consumption corresponding to each candidate detection parameter.

[0090] In this embodiment, after obtaining multiple level matching degrees corresponding to the current event, the controller determines whether the multiple level matching degrees meet preset level determination conditions. The preset level determination conditions are used to determine whether the multiple level matching degrees corresponding to the current event can clearly point to an event processing level.

[0091] Specifically, the controller sorts multiple matching scores from highest to lowest, and determines the highest and second-highest matching scores. When the highest matching score does not reach the first matching threshold, or the difference between the highest and second-highest matching scores does not reach the second matching threshold, the controller determines that multiple matching scores do not meet the preset level determination conditions.

[0092] The maximum matching degree did not reach the first matching threshold, indicating that the overall similarity between the current event and the historical dual-channel echo features corresponding to each event processing level is low; the difference between the maximum matching degree and the second largest matching degree did not reach the second matching threshold, indicating that the current event may be similar to at least two event processing levels at the same time, and the event processing level corresponding to the current event cannot be reliably determined based on the initial detection results.

[0093] In step S310, the preset candidate criteria are used to select the possible event processing level corresponding to the current event from multiple event processing levels. The preset candidate criteria are that the level matching degree corresponding to the event processing level reaches a preset candidate matching threshold.

[0094] In one implementation, the controller determines the event processing level that meets a preset candidate matching threshold as the event processing level to be distinguished. When there are fewer than two event processing levels that meet the preset candidate matching threshold, the controller determines the event processing level corresponding to the highest level matching degree and the event processing level corresponding to the second highest level matching degree as the event processing level to be distinguished.

[0095] In another implementation, the controller first determines the maximum matching degree and then determines the difference between the matching degree of each of the other matching degrees and the maximum matching degree. The controller then determines the event processing level whose corresponding difference does not exceed a preset candidate difference as the event processing level to be distinguished.

[0096] For example, the current event has a matching degree of 86%, 75%, and 30% for the first, second, and third event processing levels, respectively. The first matching threshold is 80%, and the second matching threshold is 15%. Although the maximum matching degree of 86% reaches the first matching threshold, the difference between the maximum matching degree and the second maximum matching degree is 11%, which does not reach the second matching threshold. Therefore, the controller determines that multiple matching degrees do not meet the preset level determination conditions and determines the first and second event processing levels as event processing levels to be distinguished.

[0097] In step S320, the multiple candidate detection parameters are detection parameters that are preset during the camera's factory calibration stage. At least one parameter of each candidate detection parameter differs.

[0098] The controller sequentially reads multiple candidate detection parameters and filters the corresponding historical echo feature records from the event echo feature library based on each candidate detection parameter. Specifically, the controller determines the historical dual-channel echo features in the event echo feature library that are the same as or similar to the current candidate detection parameter as the historical dual-channel echo features corresponding to that candidate detection parameter.

[0099] The controller further classifies the selected historical dual-channel echo features according to the event processing level associated with the historical dual-channel echo features, to obtain the historical echo feature set corresponding to each event processing level to be distinguished. For example, when the event processing levels to be distinguished include a first event processing level and a second event processing level, for the first candidate detection parameter, the controller obtains multiple historical dual-channel echo features corresponding to the first candidate detection parameter and associated with the first event processing level from the event echo feature library to form a first historical echo feature set; and obtains multiple historical dual-channel echo features corresponding to the first candidate detection parameter and associated with the second event processing level to form a second historical echo feature set.

[0100] The controller obtains the historical echo feature sets corresponding to each event processing level to be distinguished under the second and third candidate detection parameters in the same way.

[0101] In step S330, for each candidate detection parameter, the controller performs statistical processing on the historical echo feature set corresponding to each event processing level to be distinguished, and determines the representative feature values ​​of channel-related features such as echo arrival time difference, echo amplitude ratio and echo energy ratio in each historical echo feature set.

[0102] The representative feature value is the average value of the channel-related features of the same type in the same historical echo feature set, or it can be set as the median value of the channel-related features of the same type after sorting the values ​​from high to low.

[0103] The controller compares the representative feature values ​​corresponding to different event processing levels to determine the differences in representative feature values ​​for each type of channel-related feature. Following a preset normalization rule, the controller transforms the differences in representative feature values ​​of different types to the same value range and performs statistical processing according to preset feature weights to obtain the feature discrimination of the corresponding candidate detection parameters. The greater the difference in representative feature values, the higher the corresponding feature discrimination.

[0104] Feature discriminability is used to represent the degree of distinguishability between dual-channel echo features corresponding to different event processing levels when using the corresponding candidate detection parameters. The higher the feature discriminability, the more effective the dual-channel echo features obtained using the candidate detection parameters are in distinguishing different event processing levels; the lower the feature discriminability, the more similar the dual-channel echo features obtained for different event processing levels under the candidate detection parameters.

[0105] When there are two event processing levels to be distinguished, the controller determines the feature discrimination degree corresponding to the candidate detection parameter based on the feature distribution difference between the two historical echo feature sets. When there are three or more event processing levels to be distinguished, the controller determines the feature distribution difference between the historical echo feature sets corresponding to any two event processing levels to be distinguished, and determines the minimum value among the multiple feature distribution differences as the feature discrimination degree corresponding to the candidate detection parameter. This reduces the likelihood that the candidate detection parameter can only distinguish some event processing levels and cannot distinguish others.

[0106] In step S340, during the camera's factory calibration stage, the ultrasonic transmitter, the first ultrasonic receiver, and the second ultrasonic receiver are controlled to perform detection according to multiple candidate detection parameters, and the power consumption corresponding to each candidate detection parameter is recorded to establish the correspondence between the candidate detection parameters and the detection power consumption.

[0107] The controller associates and stores candidate detection parameters with their corresponding detection energy consumption. When it is necessary to determine the target detection parameter, the controller reads the detection energy consumption corresponding to the candidate detection parameter based on the parameter identifier of the candidate detection parameter.

[0108] In other implementations, the controller can determine the detection energy consumption corresponding to the candidate detection parameters based on at least one of the following: transmit power, number of transmitted pulses, duration of a single pulse, receiver channel on-time, and sampling frequency. Generally, higher transmit power, more transmitted pulses, or longer receiver channel on-time result in higher detection energy consumption.

[0109] Specifically, the controller first determines the candidate detection parameters whose feature discrimination reaches a preset discrimination threshold as optional detection parameters, then selects the candidate detection parameter with the lowest detection energy consumption from multiple optional detection parameters, and determines the selected candidate detection parameter as the target detection parameter.

[0110] When multiple selectable detection parameters have the same detection energy consumption, the controller determines the selectable detection parameter with the highest feature discrimination as the target detection parameter. When there is no candidate detection parameter with a feature discrimination that reaches the preset discrimination threshold, the controller determines the candidate detection parameter with the highest feature discrimination as the target detection parameter, so as to prioritize improving the ability of supplementary detection to distinguish the processing level of events to be distinguished.

[0111] For example, the feature discrimination quotients corresponding to the first candidate detection parameter, the second candidate detection parameter, and the third candidate detection parameter are 55%, 82%, and 90%, respectively, and their corresponding detection energy consumption is low, medium, and high, respectively. When the preset discrimination threshold is 80%, the controller determines the second candidate detection parameter and the third candidate detection parameter as optional detection parameters, and selects the second candidate detection parameter with lower detection energy consumption as the target detection parameter.

[0112] Therefore, in this embodiment, when multiple matching degrees cannot clearly determine the event processing level corresponding to the current event, at least two event processing levels that need to be distinguished are first determined. Then, the target detection parameters are determined based on the feature distinguishability of the event processing levels to be distinguished and the detection energy consumption of different candidate detection parameters. This achieves the goal of prioritizing candidate detection parameters with lower detection energy consumption while meeting the requirements for distinguishing the event processing levels to be distinguished, thus balancing the event distinguishability of supplementary detection with the low power consumption requirements of the camera.

[0113] It should be noted that in this embodiment, the first matching threshold, the second matching threshold, the preset candidate matching threshold, and the preset discrimination threshold are all predetermined during the camera's factory calibration stage, based on multiple calibration events with known reference event processing levels. The reference event processing level is determined based on the historical event image recognition results corresponding to the calibration event and the preset event processing strategy.

[0114] Specifically, the controller performs level matching on multiple calibration events based on the event echo feature library, and records the maximum level matching degree, the second largest level matching degree, and the reference event processing level for each calibration event. A first matching threshold is determined based on the range of the maximum level matching degree corresponding to the correct calibration event, and a second matching threshold is determined based on the difference range between the maximum and second largest level matching degrees.

[0115] For calibrated events where multiple matching degrees cannot definitively determine the event handling level, the controller determines a preset candidate matching threshold based on the range of matching degrees that allows the reference event handling level to remain as the event handling level to be distinguished. The preset candidate matching threshold is less than or equal to a first matching threshold.

[0116] The controller also performs supplementary detection on the calibration event according to multiple candidate detection parameters, and determines the preset discrimination threshold based on the feature distribution differences between the historical echo feature sets corresponding to the event processing level under different candidate detection parameters, as well as the event processing level determination results after supplementary detection.

[0117] For example, during the camera's factory calibration process, calibration results show that when the maximum level matching degree reaches 80%, and the difference between the maximum level matching degree and the second-highest level matching degree reaches 15%, the corresponding event processing level can be correctly determined. Therefore, the first matching threshold is set to 80%, and the second matching threshold is set to 15%. For calibration events where the event processing level cannot be clearly determined, when the level matching degree reaches 60%, the corresponding reference event processing level can be retained as the event processing level to be distinguished. Therefore, the preset candidate matching threshold is set to 60%. When the feature discrimination degree corresponding to the candidate detection parameter reaches 80%, the supplementary detection result can meet the preset level determination requirements. Therefore, the preset discrimination degree threshold is set to 80%.

[0118] In one embodiment, step S300 involves performing supplementary detection based on the target detection parameters to obtain the second dual-channel echo feature corresponding to the current event, including: Step S350: Control the ultrasonic transmitter to emit a second ultrasonic detection signal according to the target detection parameters; Step S360: Receive the echo corresponding to the second ultrasonic detection signal through the first ultrasonic receiver to obtain the first supplementary echo signal corresponding to the current event; Step S370: Receive the echo corresponding to the second ultrasonic detection signal through the second ultrasonic receiver to obtain the second supplementary echo signal corresponding to the current event; Step S380: Determine the second dual-channel echo feature corresponding to the current event based on at least two channel association features between the first supplementary echo signal and the second supplementary echo signal.

[0119] In this embodiment, after determining the target detection parameters, the controller generates a second transmission control command according to the target detection parameters and sends the second transmission control command to the ultrasonic transmitter to control the ultrasonic transmitter to transmit a second ultrasonic detection signal.

[0120] The target detection parameters are different from at least one of the first detection parameters, so that the second ultrasonic detection signal can obtain echo information different from that of the first ultrasonic detection signal.

[0121] The controller records the corresponding start time of the second ultrasonic detection signal and uses this start time as the common time reference for the first and second ultrasonic receivers to receive the supplementary echo.

[0122] In step S360, the first ultrasonic receiver receives the echo formed by the reflection of the second ultrasonic detection signal from the target corresponding to the current event, and obtains a first supplementary original echo signal. The first supplementary original echo signal is sampled and processed to obtain a first supplementary echo signal. The method for obtaining the second supplementary echo signal is the same as that for the first supplementary echo signal, and will not be described again here.

[0123] Steps S360 and S370 are used to distinguish between the two supplementary echo receiving channels, and do not indicate that the first supplementary echo signal and the second supplementary echo signal are obtained in chronological order. The first ultrasonic receiver and the second ultrasonic receiver receive the echoes synchronously so that the first supplementary echo signal and the second supplementary echo signal correspond to the same supplementary detection.

[0124] In step S380, the controller extracts one or more supplementary echo peaks from the first supplementary echo signal and the second supplementary echo signal, respectively. When a supplementary echo signal includes multiple supplementary echo peaks, the controller determines the supplementary echo peak with the largest amplitude as the supplementary target echo peak corresponding to that supplementary echo signal.

[0125] The controller pairs the first supplementary target echo peak corresponding to the first supplementary echo signal with the second supplementary target echo peak corresponding to the second supplementary echo signal to form a supplementary target echo peak pair, and determines at least two channel association characteristics among echo arrival time difference, echo amplitude ratio, and echo energy ratio based on the supplementary target echo peak pair. The selection method for the supplementary echo peaks is the same as the selection method for the echo peaks corresponding to the first dual-channel echo characteristics.

[0126] The controller normalizes the determined channel association features according to the same normalization rule as the first dual-channel echo feature, and arranges the at least two normalized channel association features in a preset order to form the second dual-channel echo feature corresponding to the current event.

[0127] When neither the first nor the second supplementary echo signal yields a supplementary echo peak, the controller marks the current supplementary detection result as having no valid echo. When the current supplementary detection result is marked as having no valid echo, the controller does not execute steps S410 to S450, but instead determines the preset fallback event processing level as the target event processing level corresponding to the current event, and controls the image acquisition module according to the camera control method corresponding to the preset fallback event processing level. In this embodiment, the preset fallback event processing level is specifically the event processing level for controlling the image acquisition module to perform low frame rate or short-duration image acquisition.

[0128] The second dual-channel echo feature uses the same feature type and arrangement order as the first dual-channel echo feature, so that the target event processing level corresponding to the current event can be determined jointly based on the first dual-channel echo feature and the second dual-channel echo feature.

[0129] This embodiment obtains the second dual-channel echo characteristics corresponding to the current event by performing supplementary detection, providing supplementary echo information for subsequent elimination of matching ambiguities between different event processing levels.

[0130] In one embodiment, in step S400, after obtaining the second dual-channel echo feature, determining the target event processing level corresponding to the current event based on the first dual-channel echo feature and the second dual-channel echo feature includes: Step S410: Based on the target detection parameters, determine the historical dual-channel echo features in the event echo feature library that satisfy the preset parameter correspondence conditions between the historical detection parameters associated with the target detection parameters and the historical dual-channel echo features associated with the target detection parameters as supplementary candidate dual-channel echo features; Step S420: Compare the second dual-channel echo feature with each of the supplementary candidate dual-channel echo features to determine the supplementary feature matching degree between the second dual-channel echo feature and each of the supplementary candidate dual-channel echo features; Step S430: According to the event processing level associated with each of the supplementary candidate dual-channel echo features, classify and statistically process the matching degree of multiple supplementary features to determine the supplementary level matching degree of the current event corresponding to each of the event processing levels; Step S440: The matching degree of each level determined according to the first dual-channel echo feature is fused with the supplementary level matching degree of the corresponding event processing level determined according to the second dual-channel echo feature to obtain the updated level matching degree of the current event corresponding to each event processing level. Step S450: Determine the target level matching degree from the multiple update level matching degrees, and determine the event processing level corresponding to the target level matching degree as the target event processing level corresponding to the current event.

[0131] In this embodiment, in step S410, the controller filters multiple historical echo feature records in the event echo feature library according to the target detection parameters. The historical detection parameters in each historical echo feature record are compared with the target detection parameters, and historical dual-channel echo features whose associated historical detection parameters and target detection parameters meet the preset parameter correspondence conditions are determined as supplementary candidate dual-channel echo features.

[0132] The preset parameter correspondence conditions and judgment methods between the target detection parameters and historical detection parameters are the same as those between the first detection parameters and historical detection parameters in step S210, and will not be repeated here.

[0133] In step S420, the controller compares the channel-related features in the second dual-channel echo features with the corresponding channel-related features in the supplementary candidate dual-channel echo features according to the feature arrangement order, and determines the feature differences between the channel-related features.

[0134] The controller determines the sub-item matching degree corresponding to each channel's associated feature based on the pre-stored correspondence between feature difference ranges and sub-item matching degrees, and performs statistical processing on multiple sub-item matching degrees according to preset feature weights to obtain the supplementary feature matching degree between the second dual-channel echo feature and the corresponding supplementary candidate dual-channel echo feature. The specific method for determining the supplementary feature matching degree is the same as the method for determining the feature matching degree in step S220, and will not be repeated here.

[0135] In step S430, the controller reads the event processing level associated with each supplementary candidate dual-channel echo feature, and assigns the supplementary feature matching degree corresponding to the supplementary candidate dual-channel echo features associated with the same event processing level to the same supplementary feature matching degree set.

[0136] The controller averages the multiple supplementary feature matching scores in each supplementary feature matching score set, and determines the average result as the supplementary level matching score of the current event corresponding to each event processing level. Thus, multiple supplementary level matching scores are obtained for the current event corresponding to multiple event processing levels.

[0137] In step S440, the multiple level matching degrees corresponding to the first dual-channel echo feature and the multiple supplementary level matching degrees corresponding to the second dual-channel echo feature adopt the same value range. The controller fuses the level matching degrees and supplementary level matching degrees corresponding to the same event processing level according to the event processing level.

[0138] In one implementation, the controller averages the level matching degree and the supplementary level matching degree corresponding to the same event handling level, and determines the average result as the updated level matching degree corresponding to the event handling level.

[0139] In other embodiments, the controller sets corresponding fusion weights for the level matching degree and the supplementary level matching degree based on the feature discrimination degrees corresponding to the first detection parameter and the target detection parameter, respectively. The higher the feature discrimination degree corresponding to the detection parameter, the higher the fusion weight corresponding to the level matching degree determined by the dual-channel echo features obtained from the detection parameter. For example, the level matching degrees of the first dual-channel echo features corresponding to the first event processing level, the second event processing level, and the third event processing level are 86%, 75%, and 30%, respectively; the supplementary level matching degrees of the second dual-channel echo features corresponding to the three event processing levels are 70%, 92%, and 28%, respectively. The controller averages the level matching degree and the supplementary level matching degree corresponding to the same event processing level, respectively, to obtain updated level matching degrees of 78%, 83.5%, and 29% for the three event processing levels.

[0140] In step S450, the controller sorts the multiple update level matching degrees in descending order of matching degree, determines the update level matching degree with the largest value as the target level matching degree, and determines the event processing level corresponding to the target level matching degree as the target event processing level corresponding to the current event.

[0141] Taking the example above, the update level matching degree corresponding to the second event handling level is 83.5%, which is higher than the update level matching degree corresponding to the first and third event handling levels. Therefore, the controller determines the second event handling level as the target event handling level corresponding to the current event.

[0142] Therefore, this embodiment integrates the level matching degree determined based on the first dual-channel echo characteristics with the supplementary level matching degree determined based on the second dual-channel echo characteristics, thereby making comprehensive use of the echo information obtained from the initial detection and supplementary detection, and reducing the impact of accidental differences in single detection results on the event processing level determination result.

[0143] In one embodiment, step S400, controlling the image acquisition module to maintain a low-power state or switch from a low-power state to an image acquisition state according to the target event processing level, includes: Step S460: Determine the target camera control method corresponding to the target event processing level according to the preset event processing strategy; Step S470: When the target camera control mode is the first control mode, a low power hold command is generated, and the image acquisition module is controlled to maintain the low power state according to the low power hold command; Step S480: When the target camera control mode is the second control mode, a state switching instruction is generated, and the image acquisition module is controlled to switch from the low power state to the image acquisition state according to the state switching instruction, so as to acquire the event image corresponding to the current event.

[0144] In this embodiment, in step S460, the controller obtains the correspondence between the event processing level and the camera control mode from the preset event processing strategy, and determines the target camera control mode corresponding to the target event processing level based on the correspondence.

[0145] The preset event handling strategy pre-stores the camera control methods corresponding to each event handling level. The target camera control methods include a first control method for keeping the image acquisition module in a low-power state, and a second control method for switching the image acquisition module to the image acquisition state.

[0146] In step S470, when the target camera control mode is the first control mode, the controller generates a low-power hold command and sends the low-power hold command to the image acquisition module. Based on the low-power hold command, the image acquisition module controls the image sensor, image signal processor, or at least some circuitry related to image acquisition to remain in a low-power state. The low-power state includes, but is not limited to, sleep mode, power-off mode, low-frequency operation, or stopped image output.

[0147] The first control method is suitable for target event processing levels that do not require event image acquisition according to the preset event processing strategy. By controlling the image acquisition module to continue to maintain a low-power state, it is possible to avoid the current event causing the image acquisition module to wake up ineffectively.

[0148] In step S480, when the target camera control mode is the second control mode, the controller generates a state switching command and sends the state switching command to the image acquisition module. Based on the state switching command, the image acquisition module controls the image sensor and image signal processor to switch from a low-power state to a working state to acquire the event image corresponding to the current event.

[0149] The controller also associates and stores the acquired event images with the event identifiers corresponding to the current event, so that the event can be viewed or processed later. After the event image acquisition for the current event is completed, the controller can switch the image acquisition module back to a low-power state.

[0150] This embodiment selectively controls the image acquisition module to maintain a low-power state or switch to image acquisition state based on the target event processing level, so that events that do not require image acquisition will not wake up the image acquisition module, thereby reducing invalid startup of the image acquisition module and invalid image acquisition.

[0151] In one embodiment, the method further includes: Step S250: Sort the multiple matching degrees in descending order of matching degree to determine the first matching degree and the second matching degree, wherein the first matching degree is the maximum value among the multiple matching degrees and the second matching degree is the second largest value among the multiple matching degrees. Step S260: Compare the first level matching degree with the first matching threshold, and compare the matching degree difference between the first level matching degree and the second level matching degree with the second matching threshold; Step S270: When the first level matching degree reaches the first matching threshold and the matching degree difference reaches the second matching threshold, determine that multiple level matching degrees meet the preset level determination conditions, and determine the event processing level corresponding to the first level matching degree as the target event processing level corresponding to the current event; Step S280: According to the camera control mode corresponding to the target event processing level, control the image acquisition module to maintain the low power state or switch from the low power state to the image acquisition state.

[0152] In this embodiment, when the multiple level matching degrees determined in step S200 meet the preset level determination conditions, the controller does not perform supplementary detection, but directly determines the target event processing level corresponding to the current event based on the multiple level matching degrees.

[0153] In step S250, the controller sorts the multiple matching degrees in descending order of matching degree, determines the maximum value among the multiple matching degrees as the first level matching degree, and determines the second maximum value among the multiple matching degrees as the second level matching degree.

[0154] In step S260, the controller compares the first-level matching degree with the first matching threshold and determines the matching degree difference between the first-level matching degree and the second-level matching degree. The controller then compares this matching degree difference with the second matching threshold.

[0155] The first matching threshold and the second matching threshold are thresholds determined and stored in the memory during the aforementioned camera factory calibration stage. The specific determination method has been explained in the aforementioned embodiments and will not be repeated here.

[0156] In step S270, when the first level matching degree reaches the first matching threshold and the matching degree difference between the first level matching degree and the second level matching degree reaches the second matching threshold, the controller determines that multiple level matching degrees meet the preset level determination conditions.

[0157] The controller determines the first-level matching degree as the target-level matching degree, and determines the event handling level corresponding to the first-level matching degree as the target event handling level corresponding to the current event.

[0158] For example, the matching degree of the current event to the first event processing level, the second event processing level, and the third event processing level are 90%, 70%, and 35%, respectively, with a first matching threshold of 80% and a second matching threshold of 15%. Specifically, the first level matching degree is 90%, reaching the first matching threshold; the matching degree difference between the first and second levels is 20%, reaching the second matching threshold. Therefore, the controller determines that multiple level matching degrees meet the preset level determination conditions and determines the first event processing level as the target event processing level corresponding to the current event.

[0159] In step S280, the controller determines the camera control mode corresponding to the target event processing level according to the preset event processing strategy, and controls the image acquisition module to maintain a low-power state or switch from a low-power state to an image acquisition state according to the camera control mode. The specific control mode is the same as the control mode described in steps S460 to S480, and will not be repeated here.

[0160] This embodiment makes a judgment based on the multiple matching degrees obtained in the initial detection. If the event processing level corresponding to the current event can be clearly determined, the image acquisition module is directly controlled according to the multiple matching degrees, and no further detection is performed, thereby reducing unnecessary ultrasonic detection times and detection energy consumption.

[0161] In one embodiment, such as Figure 2 As shown, a camera event triggering device is also provided, applied to a camera, the camera including an image acquisition module, an ultrasonic transmitter, a first ultrasonic receiver, and a second ultrasonic receiver, the device including: The echo feature determination module is used to control the ultrasonic transmitter to emit a first ultrasonic detection signal when the image acquisition module is in a low power state, and to determine the first dual-channel echo feature of the current event based on the echo signals received by the first ultrasonic receiver and the second ultrasonic receiver. The level matching determination module is used to perform level matching on the first dual-channel echo features based on a pre-established event echo feature library, and determine multiple level matching degrees corresponding to the current event, wherein each level matching degree corresponds to an event processing level. The supplementary detection execution module is used to determine the target detection parameters when multiple level matching degrees do not meet the preset level determination conditions, and to perform supplementary detection based on the target detection parameters to obtain the second dual-channel echo feature corresponding to the current event. The power consumption state control module is used to determine the target event processing level corresponding to the current event based on the first dual-channel echo feature and the second dual-channel echo feature after obtaining the second dual-channel echo feature, and control the image acquisition module to maintain a low power consumption state or switch from a low power consumption state to an image acquisition state based on the target event processing level.

[0162] It should be noted that the specific processing procedures of the echo feature determination module, level matching determination module, supplementary detection execution module, and power consumption state control module described in this embodiment correspond to the processes in the aforementioned method embodiments. The specific limitations in the aforementioned method embodiments are all applicable to the device described in this embodiment and executed by the corresponding modules, and will not be repeated here. The modules described in this embodiment can be implemented by a processor executing program instructions stored in memory, or they can be implemented by software, hardware, or a combination of both, depending on actual needs. The device described in this embodiment can implement the processing procedures described in the aforementioned method embodiments and achieve the same or corresponding technical effects as the aforementioned method embodiments.

[0163] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0164] One embodiment of this application also provides a computer device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the steps in any of the above-described methods.

[0165] The embodiments described above are merely examples of several implementation methods of this application. They are described in a relatively specific and detailed manner, but should not be construed as limiting the scope of the invention.

Claims

1. A method for triggering camera events, characterized in that, Applied to a camera, the camera including an image acquisition module, an ultrasonic transmitter, a first ultrasonic receiver, and a second ultrasonic receiver, the method includes: When the image acquisition module is in a low-power state, the ultrasonic transmitter is controlled to emit a first ultrasonic detection signal, and the first dual-channel echo feature of the current event is determined based on the echo signals received by the first ultrasonic receiver and the second ultrasonic receiver. Based on a pre-established event echo feature library, the first dual-channel echo features are matched for level to determine multiple level matching degrees corresponding to the current event, wherein each level matching degree corresponds to an event processing level. When multiple matching degrees of the stated level do not meet the preset level determination conditions, the target detection parameters are determined, and supplementary detection is performed based on the target detection parameters to obtain the second dual-channel echo feature corresponding to the current event. After obtaining the second dual-channel echo feature, the target event processing level corresponding to the current event is determined based on the first dual-channel echo feature and the second dual-channel echo feature, and the image acquisition module is controlled to maintain a low-power state or switch from a low-power state to an image acquisition state based on the target event processing level. The event echo feature library is pre-established, including: Acquire historical detection parameters, historical first echo signals, historical second echo signals, and historical event images corresponding to multiple historical events; Based on at least two channel association characteristics between the historical first echo signal and the historical second echo signal corresponding to the same historical event, the historical dual-channel echo characteristics corresponding to the historical event are determined. Target recognition is performed on the historical event images to obtain the historical image recognition results corresponding to the historical events; Based on the historical image recognition results and the preset event processing strategy, the event processing level corresponding to the historical event is determined, wherein different event processing levels correspond to different camera control methods; The historical detection parameters, historical dual-channel echo characteristics, and event processing levels corresponding to the same historical event are associated and stored to establish an event echo feature library. Based on a pre-established event echo feature library, the first dual-channel echo features are subjected to level matching to determine multiple level matching degrees corresponding to the current event, including: Based on the first detection parameter corresponding to the first ultrasonic detection signal, the historical dual-channel echo features associated with the event echo feature library and the first detection parameter that meet the preset parameter correspondence conditions are determined as candidate historical dual-channel echo features. The first dual-channel echo feature is compared with each of the candidate historical dual-channel echo features to determine the feature matching degree between the first dual-channel echo feature and each of the candidate historical dual-channel echo features. According to the event processing level associated with each candidate historical dual-channel echo feature, the multiple feature matching degrees are classified to obtain multiple feature matching degree sets, wherein each feature matching degree set corresponds to an event processing level. According to preset statistical rules, the feature matching degrees in multiple feature matching degree sets are statistically processed to determine the level matching degree of the current event corresponding to each of the event processing levels; After obtaining the second dual-channel echo feature, the target event processing level corresponding to the current event is determined based on the first dual-channel echo feature and the second dual-channel echo feature, including: Based on the target detection parameters, historical dual-channel echo features that satisfy the preset parameter correspondence conditions between the historical detection parameters associated with the event echo feature library and the target detection parameters are identified as supplementary candidate dual-channel echo features. The second dual-channel echo feature is compared with each of the supplementary candidate dual-channel echo features to determine the supplementary feature matching degree between the second dual-channel echo feature and each of the supplementary candidate dual-channel echo features; According to the event processing level associated with each of the supplementary candidate dual-channel echo features, the matching degree of multiple supplementary features is classified and statistically processed to determine the supplementary level matching degree of the current event corresponding to each of the event processing levels; The matching degree of each level determined according to the first dual-channel echo feature is fused with the supplementary level matching degree of the corresponding event processing level determined according to the second dual-channel echo feature to obtain the updated level matching degree of the current event corresponding to each event processing level. A target level matching degree is determined from multiple update level matching degrees, and the event processing level corresponding to the target level matching degree is determined as the target event processing level corresponding to the current event.

2. The camera event triggering method according to claim 1, characterized in that, Controlling the ultrasonic transmitter to emit a first ultrasonic detection signal, and determining a first dual-channel echo characteristic of the current event based on the echo signals received by the first and second ultrasonic receivers, including: The ultrasonic transmitter is controlled to emit a first ultrasonic detection signal according to the first detection parameters. The first echo signal corresponding to the current event is obtained through the first ultrasonic receiver; The second echo signal corresponding to the current event is obtained through the second ultrasonic receiver; The first dual-channel echo feature corresponding to the current event is determined based on at least two channel association features between the first echo signal and the second echo signal.

3. The camera event triggering method according to claim 1, characterized in that, When multiple matching degrees of the stated level do not meet the preset level determination conditions, the target detection parameters are determined, including: When multiple level matching degrees do not meet the preset level determination conditions, at least two event processing levels that meet the preset candidate conditions are determined as event processing levels to be distinguished based on the multiple level matching degrees. For each candidate detection parameter among multiple candidate detection parameters, historical dual-channel echo features corresponding to the candidate detection parameter and respectively associated with each of the event processing levels to be distinguished are obtained from the event echo feature library to obtain a set of historical echo features corresponding to each of the event processing levels to be distinguished. Representative feature values ​​of at least two channel-related features in each of the historical echo feature sets are determined respectively, and feature discrimination degree corresponding to each candidate detection parameter is determined based on the difference between the representative feature values ​​corresponding to different event processing levels to be distinguished. The detection energy consumption corresponding to each candidate detection parameter is determined, and the target detection parameter is determined from the multiple candidate detection parameters based on the feature discrimination and detection energy consumption corresponding to each candidate detection parameter.

4. The camera event triggering method according to claim 3, characterized in that, Supplementary detection is performed based on the target detection parameters to obtain the second dual-channel echo feature corresponding to the current event, including: The ultrasonic transmitter is controlled to emit a second ultrasonic detection signal according to the target detection parameters; The first supplementary echo signal corresponding to the current event is obtained by receiving the echo corresponding to the second ultrasonic detection signal through the first ultrasonic receiver. The second ultrasonic receiver receives the echo corresponding to the second ultrasonic detection signal to obtain the second supplementary echo signal corresponding to the current event. The second dual-channel echo feature corresponding to the current event is determined based on at least two channel association features between the first supplementary echo signal and the second supplementary echo signal.

5. A camera event triggering device applying the camera event triggering method according to any one of claims 1 to 4, characterized in that, Applied to a camera, the camera includes an image acquisition module, an ultrasonic transmitter, a first ultrasonic receiver, and a second ultrasonic receiver; the device includes: The echo feature determination module is used to control the ultrasonic transmitter to emit a first ultrasonic detection signal when the image acquisition module is in a low power state, and to determine the first dual-channel echo feature of the current event based on the echo signals received by the first ultrasonic receiver and the second ultrasonic receiver. The level matching determination module is used to perform level matching on the first dual-channel echo features based on a pre-established event echo feature library, and determine multiple level matching degrees corresponding to the current event, wherein each level matching degree corresponds to an event processing level. The supplementary detection execution module is used to determine the target detection parameters when multiple level matching degrees do not meet the preset level determination conditions, and to perform supplementary detection based on the target detection parameters to obtain the second dual-channel echo feature corresponding to the current event. The power consumption state control module is used to determine the target event processing level corresponding to the current event based on the first dual-channel echo feature and the second dual-channel echo feature after obtaining the second dual-channel echo feature, and control the image acquisition module to maintain a low power consumption state or switch from a low power consumption state to an image acquisition state based on the target event processing level.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

Images