Self-adaptive video recording method, control device and hunting camera

Abstract

The invention discloses a self-adaptive video recording method, a control device and a hunting camera, and relates to the technical field of cameras, the self-adaptive video recording method is applied to the hunting camera, and the method comprises the following steps: under the condition that a passive infrared sensor triggers a detection signal, the hunting camera is controlled to start basic video recording, and the basic video recording has an initial target recording duration; continuously acquiring a video picture shot in a video recording process, and identifying an object of the acquired video picture to obtain an object identification result signal; when it is determined that the target animal category does not exist in the video picture based on the object recognition result signal, video recording is stopped after the current basic video recording is finished; and when it is determined that the target animal category exists in the video picture based on the object recognition result signal and the time difference between the current recording time length of the hunting camera and the target recording time length is smaller than the first time difference, the target recording time length is prolonged. The invention aims to improve the hunting video recording capability of the hunting camera so as to improve the content effectiveness of the picture shot by the hunting camera.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of camera technology, in particular to an adaptive video recording method, a control device and a hunting camera. BACKGROUND

[0002] In the field of wildlife observation, the trigger recording mode of the traditional hunting camera is difficult to record the continuity of the animal behavior completely. For example, the behavior fragments with research value such as foraging process and group interaction are often truncated, resulting in incomplete observation data. Secondly, when capturing a short-term appearing animal, the camera continuously records a large number of meaningless empty pictures, which seriously shortens the continuous working time of the device in the wild. Therefore, the hunting performance of the traditional hunting camera is weak, and the picture efficiency is low. SUMMARY

[0003] The main purpose of the present application is to provide an adaptive video recording method, a control device and a hunting camera, which aims to improve the hunting recording ability of the hunting camera to improve the content effectiveness of the pictures taken by the camera.

[0004] In order to achieve the above purpose, the adaptive video recording method provided by the present application is applied to a hunting camera, the hunting camera includes a passive infrared sensor, and the adaptive video recording method includes:

[0005] In the case that the passive infrared sensor triggers a detection signal, the hunting camera is controlled to start basic video recording, and the basic video recording has an initial target recording time length;

[0006] In the case that the hunting camera is in a video recording state, the recording pictures taken by the hunting camera in the process of video recording are continuously acquired, and the objects of the acquired recording pictures are identified to obtain an object identification result signal;

[0007] In the case that it is determined based on the object identification result signal that there is no target animal category in the recording pictures, the video recording is stopped after the current basic video recording is ended;

[0008] In the case that it is determined based on the object identification result signal that there is a target animal category in the recording pictures, and the time difference between the current recording time length of the hunting camera and the target recording time length is less than a first time difference, the target recording time length is extended.

[0009] In an embodiment, the continuously acquiring the recording pictures taken by the hunting camera in the process of basic video recording of the hunting camera in the case that the hunting camera is in a video recording state includes:

[0010] Acquiring a picture capture period;

[0011] In the case that the hunting camera is in a video recording state, the recording pictures taken by the hunting camera in the process of basic video recording are acquired according to the picture capture period.

[0012] In an embodiment, the adaptive recording method further comprises:

[0013] configuring an initial target recording duration as a first preset duration;

[0014] In a case where it is determined based on the object recognition result signal that the target animal category exists in the recording picture, and a time difference between the current recording duration of the hunting camera and the target recording duration is less than a first time difference, the target recording duration is extended, and the extending the target recording duration comprises:

[0015] In a case where it is determined based on the object recognition result signal that the target animal category exists in the recording picture, and a time difference between the current recording duration of the hunting camera and the target recording duration is less than a first time difference, the target recording duration is extended by a second preset duration, and the first preset duration is greater than the second preset duration.

[0016] In an embodiment, the first preset duration is not less than three seconds and not greater than ten seconds.

[0017] In an embodiment, the identifying the obtained recording picture to obtain the object recognition result signal comprises:

[0018] inputting the obtained recording picture into a preset recognition model;

[0019] performing animal image detection on the recording picture by the preset recognition model, and outputting a detection result;

[0020] determining the object recognition result signal based on the detection result.

[0021] In an embodiment, the detection result comprises an animal category and a corresponding confidence; and the determining the object recognition result signal based on the detection result comprises:

[0022] in a case where it is determined that the animal category in the detection result is a target animal category and the corresponding confidence is greater than a preset value, it is determined that the target animal category exists in the recording picture;

[0023] in a case where it is determined that the animal category in the detection result is a target animal category and the corresponding confidence is not greater than a preset value, it is determined that the target animal category does not exist in the recording picture;

[0024] in a case where it is determined that the animal category in the detection result is not a target animal category, it is determined that the target animal category does not exist in the recording picture.

[0025] In an embodiment, the adaptive recording method further comprises:

[0026] When the hunting camera is in recording mode, it continuously acquires the video footage captured during the recording process and identifies the positions of objects in the acquired video footage to obtain the position recognition result signal.

[0027] If, based on the location recognition result signal, it is determined that the object in the video footage is more than a preset distance from the hunting camera, recording will stop after the current basic recording ends.

[0028] In one embodiment, the adaptive recording method further includes:

[0029] The timer starts when the hunting camera begins recording;

[0030] Once the timer reaches the third preset duration, control the hunting camera to stop recording.

[0031] This application also provides a control device for implementing the adaptive recording method as described above.

[0032] This application also provides a hunting camera, which includes the control device as described above.

[0033] As can be seen from the above, the adaptive recording method, control device and hunting camera provided in this application dynamically control the start and stop of recording by identifying the target animal category in the recording screen in real time. When the target animal is detected, the recording time is automatically extended, and the recording is terminated in time when no target animal is detected. This not only ensures the integrity of the animal behavior record, but also effectively reduces the waste of energy and storage resources caused by invalid recording. It has the advantage of intelligently adjusting the recording time according to the appearance of the target animal to balance the integrity of the observation and the endurance of the equipment. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0035] Figure 1 A flowchart of the first embodiment of the adaptive recording method provided in this application;

[0036] Figure 2 A flowchart of the second embodiment of the adaptive recording method provided in this application;

[0037] Figure 3 A flowchart of the third embodiment of the adaptive recording method provided in this application;

[0038] Figure 4A flowchart of the fourth embodiment of the adaptive recording method provided in this application;

[0039] Figure 5 A flowchart of the fifth embodiment of the adaptive recording method provided in this application;

[0040] Figure 6 A flowchart of the sixth embodiment of the adaptive recording method provided in this application.

[0041] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0042] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0043] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the use of "and / or" or "and / or" throughout the text includes three parallel solutions. For example, "A and / or B" includes solution A, solution B, or a solution that simultaneously satisfies A and B. Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0044] It should be noted that step designations such as S100 and S200 are used in this document for the purpose of more clearly and concisely describing the corresponding content, and do not constitute a substantial limitation on the order. In specific implementation, those skilled in the art may execute S200 first and then S100, etc., but these should all be within the protection scope of this application.

[0045] In traditional infrared-triggered hunting camera systems, the passive infrared sensor initiates a fixed-duration recording mechanism upon detecting movement, leading to an irreconcilable conflict between storage space utilization and the integrity of animal behavior recordings. When environmental disturbances trigger the sensor, the fixed-duration recording mechanism continuously generates invalid video data, resulting in ineffective use of storage media. When the target animal appears, the preset recording duration cannot dynamically adapt to the animal's activity cycle; too short a duration causes breaks in the behavioral sequence, while too long a duration generates redundant video frames.

[0046] For example, in a temperate forest ecological monitoring scenario, deployed hunting cameras employ a fixed 15-second recording strategy. When wind speeds reach 5 m / s, causing shrub branches to sway, the passive infrared sensor triggers recording, generating a 15-second video file containing no animal targets; this invalid data occupies storage space. Simultaneously, when target animals such as red foxes move intermittently, the initial 15-second recording only captures their entry into the monitoring area; subsequent activity is lost due to recording termination. When tits or other birds of prey rapidly pass through the monitoring area, only the first two seconds of the complete 15-second recording contain valid targets, with the remaining time continuously consuming battery power. The system cannot analyze the footage in real-time during recording, causing all triggered events to force the completion of the preset recording duration.

[0047] If these issues are not addressed, the storage medium will become completely filled with video data triggered by non-target objects within a certain period, forcing the device into a data overwrite mode and resulting in the loss of valuable animal behavior records. Battery life will also be shortened due to continuous processing of invalid video, leading to shorter device maintenance cycles.

[0048] In response, this application proposes an adaptive recording method for use in a hunting camera, which includes a passive infrared sensor.

[0049] In one embodiment, such as Figure 1 As shown, the adaptive recording method includes steps S100 to S400.

[0050] In this embodiment, step S100 involves controlling the hunting camera to start basic recording when the passive infrared sensor triggers a detection signal.

[0051] In this embodiment, step S200 involves continuously acquiring the video footage captured during the recording process while the hunting camera is in recording mode, and identifying the objects in the acquired video footage to obtain an object identification result signal.

[0052] In this embodiment, step S300 involves stopping recording after the current basic recording ends if it is determined that there is no target animal category in the video frame based on the object recognition result signal.

[0053] In this embodiment, step S400 involves extending the target recording time if, based on the object recognition result signal, it is determined that a target animal category exists in the video recording and the time difference between the current recording time of the hunting camera and the target recording time is less than a first time difference.

[0054] Among them, the passive infrared sensor trigger detection signal refers to the detection signal being triggered by detecting changes in heat generated by animal activity using an infrared sensor. Specifically, a pyroelectric infrared sensor can be used, which generates an electrical signal by sensing changes in the temperature difference between the organism and its environment. This feature solves the problem of traditional motion detection being easily affected by environmental interference, thus improving the accuracy of the trigger signal.

[0055] The basic recording stage refers to the initial recording phase triggered by infrared radiation. This can be achieved using a fixed-duration video recording module, for example, by setting an initial target recording duration, which is not limited here. This ensures that the initial footage of the animal's appearance is fully captured at the moment of triggering, avoiding the loss of crucial footage due to response delays.

[0056] Continuously acquiring video footage and identifying objects involves periodically capturing and analyzing video frames during the basic recording process. This can be achieved by combining an image sampling frequency of once per second with an object detection algorithm. In this way, real-time video analysis provides data support for dynamically adjusting recording strategies.

[0057] The object recognition result signal can refer to recognition data that includes a determination of the existence of the target animal type, or it can refer to recognition data that includes whether or not it is an animal type. Specifically, it can be implemented using the classification results and confidence parameters output by a convolutional neural network. This feature achieves accurate animal category determination through a machine learning model, providing a reliable basis for subsequent video recording decisions.

[0058] Understandably, this embodiment integrates a passive infrared triggering mechanism with real-time image recognition technology to construct a dynamic video recording control system. A dual triggering mechanism ensures a combination of rapid response and intelligent judgment. During the triggering phase, basic recording is used to ensure the capture of key images, while during the continuous phase, real-time image analysis dynamically determines whether to continue or terminate the recording. This effectively solves the contradiction between wasted storage space and missing key data inherent in traditional fixed-duration recording, achieving an optimal balance between the completeness of animal behavior recording and equipment energy efficiency.

[0059] In one feasible implementation, the adaptive recording method is applied to a hunting camera containing a passive infrared sensor. When the passive infrared sensor triggers a detection signal, it controls the hunting camera to start basic recording. During recording, the camera continuously acquires video footage and identifies objects in the footage, obtaining an object identification result signal. Based on the object identification result signal, it is determined whether a target animal category exists in the video footage. If no target animal category exists, recording stops after the current basic recording ends (i.e., the recording duration based on the basic recording reaches the target recording duration). If a target animal category exists, and the time difference between the current recording duration and the target recording duration is less than a first time difference, it indicates that recording is nearing its end, but recording is still necessary to continuously acquire valid information. In this case, the target recording duration of the basic recording is extended. Thus, by combining passive infrared triggering and real-time image recognition, intelligent recording control is achieved.

[0060] The specific value of the first time difference can be 1 second or 2 seconds, depending on the actual application requirements, to ensure the continuity of recording. Understandably, if the first time difference is 1 second and the target recording duration is 5 seconds, if the target animal category can still be detected after 4 seconds, the target recording duration will be extended.

[0061] Extending the target recording duration for the base recording ensures that newly added video content remains continuous with the current video segment, preventing the creation of new independent video files. Instead, it is directly appended to the current recording file as a continuation of the same video stream. This continuous recording mechanism avoids video fragmentation issues caused by segmented recording, effectively guaranteeing the integrity and temporal continuity of animal behavior records. Simultaneously, this design also helps reduce the system overhead caused by frequent file creation and closure by the file system, improving storage efficiency and device stability.

[0062] Optionally, when a target animal is detected, the duration of the base video recording is automatically extended to form a continuous video sequence, ensuring complete recording of animal behavior. When no target animal is detected, recording is stopped promptly to avoid generating invalid data. This overcomes the limitations of traditional fixed-duration video recording. It can intelligently adjust the recording duration according to the actual scenario, improving storage space and energy utilization efficiency while ensuring the integrity of critical data.

[0063] In one feasible implementation, a hunting camera equipped with a passive infrared sensor and an image recognition module is deployed in a wildlife monitoring scenario. The initial target recording duration of the basic recording unit is set to 5 seconds. When the passive infrared sensor detects a moving object and triggers a detection signal, the hunting camera immediately starts a 5-second basic recording. During the recording, one frame is extracted from the recorded screen every second and input into a pre-trained animal recognition model for analysis. If any frame is identified as belonging to the target animal category (such as deer, fox, etc.) during the 5-second basic recording period, and the current recording duration reaches 4 seconds, the target recording duration is extended, for example, by 3 seconds, thus temporarily increasing the target recording duration to 8 seconds. This process continues until no target animal type is detected. If no target animal type is identified in the last second of the initial 5-second target recording duration, or in the last second of the extended 8-second target recording duration, recording stops after the current recording ends. The hunting camera returns to standby mode, waiting for the next trigger of the passive infrared sensor.

[0064] It's important to note that the initial target recording duration for basic video recording can be set to 3 or 5 seconds; it doesn't need to be too long. Otherwise, when capturing fleeting animals, such as birds, a large amount of invalid video will be generated, consuming memory.

[0065] Through the above scheme, this application realizes intelligent video recording control based on target recognition. When environmental disturbances such as swaying tree branches trigger recording, the system can stop recording promptly after one basic recording unit, avoiding the continuous generation of invalid data. When a target animal appears, the system can dynamically extend the recording time to ensure the complete capture of the animal's behavioral sequence. In this way, the adaptive control scheme improves the efficiency of storage space utilization and reduces the generation of invalid data.

[0066] Meanwhile, through real-time content analysis, the system can intelligently determine the value of video recordings, avoiding energy waste caused by fixed-duration recordings. Furthermore, the ability to dynamically adjust recording duration allows the system to better adapt to the activity patterns of different animals, improving the completeness and consistency of behavioral records.

[0067] In one embodiment, such as Figure 2 As shown, step S200 also includes steps S210 and S220.

[0068] In this embodiment, step S210 is to obtain the image capture period.

[0069] In this embodiment, step S220 involves acquiring the video footage captured by the hunting camera during basic video recording, according to the image capture cycle, while the hunting camera is in recording mode.

[0070] Optionally, the image capture period should be shorter than the initial target recording duration to ensure that at least one sample can be taken within the initial target recording duration.

[0071] The frame capture period is set to be less than the total duration of a single basic recording session. For example, if the initial target recording duration is 5 seconds, the frame capture period can be set to 0.5 to 2 seconds. The ratio between the frame capture period and the initial target recording duration ensures that multiple frames can be captured during a single recording session. For example, in a 5-second recording, at least 5 frames can be captured at a 1-second interval.

[0072] Optionally, the specific value of the image capture period can be achieved through preset parameters or dynamic adjustment, such as automatically shortening or lengthening the period according to the ambient light intensity.

[0073] Optionally, the periodically acquired footage covers the entire basic recording process, for example, sampling can be performed at the beginning, middle, and end of recording. The coordinated control of the capture period and the basic recording is achieved through a timing module, for example, by starting a periodic timer simultaneously when recording begins, triggering the image capture action.

[0074] After basic recording begins, the timing module controls the image sensor to capture frames at preset frame capture intervals. For example, if the initial target recording duration is 5 seconds and the capture interval is 1 second, one frame is extracted from the continuously recording video stream every second. The extracted frames are transmitted to the recognition module for real-time analysis. For instance, if a target animal is detected in the frame captured at the 3rd second, the system will extend the corresponding target recording duration near the end of the current basic recording. By shortening the capture interval, for example, adjusting it from 2 seconds to 0.5 seconds, the frame sampling density can be increased, allowing for continuous capture of changes in the animal's movements during rapid motion, thus avoiding missed detections due to excessively long intervals between captures.

[0075] Through the above technical solution, this application can capture multiple frames at a high frequency during basic video recording, avoiding the omission of key frames due to excessively long single capture intervals. Furthermore, by continuously acquiring frames according to a shorter frame capture cycle, the subsequent animal identification process can be analyzed based on a more dense time-series of frames, reducing the risk of misjudgment or missed judgment caused by frame update delays.

[0076] In some of the solutions described above in this application, continuous recording of animal detection is achieved by controlling the target recording duration. However, if the same preset duration is continuously used during a single detection signal trigger, it will result in wasted power for target animals that are present for extended periods. Specifically, the first trigger requires a relatively long time to ensure the capture of the full picture of animal behavior, but if the same duration is used for subsequent continuous recording, it will lead to an increase in redundant video segments, making it impossible to dynamically adjust the recording duration according to the actual scene.

[0077] In one embodiment, the adaptive recording method further includes step S510, and step S400 includes step S520.

[0078] In this embodiment, S510, the initial target recording duration is configured as a first preset duration.

[0079] In this embodiment, S520, if it is determined that there is a target animal category in the video frame based on the object recognition result signal, and the time difference between the current recording time of the hunting camera and the target recording time is less than the first time difference, the second preset time is extended based on the target recording time, and the first preset time is greater than the second preset time.

[0080] The initial target recording duration is set to cover the initial behavioral phase when an animal enters the monitored area, i.e., the first preset duration, for example, a range of three to ten seconds, to ensure complete recording of the animal's sudden running or foraging actions upon first triggering. Subsequent extended target recording durations (Nth extension) can be shortened to a smaller second preset duration, such as one to five seconds, dynamically reducing overall energy consumption by decreasing the single recording time. The difference in recording duration between the initial and subsequent recordings is achieved through a stepped adjustment mechanism, with a decreasing relationship between the two, and the second preset duration is strictly less than the first preset duration. Thus, while ensuring the integrity of the initial recording, the generation of invalid video data is reduced by gradually lowering the single recording duration threshold. Furthermore, while maintaining necessary monitoring coverage, redundant recording segments caused by fixed durations are effectively reduced, preventing excessive power consumption from continuous high-power recording operations.

[0081] In one feasible implementation, during a single detection signal trigger, a first preset duration is configured to be 10 seconds, and a second preset duration is configured to be 5 seconds. When the passive infrared sensor detects an animal for the first time, the system initiates the first basic recording, with a recording duration of 10 seconds. If the animal is continuously detected within these 10 seconds, and the time difference between the current recording duration and the target recording duration is less than the first duration, the recording duration is extended by 5 seconds. Subsequently, as long as the animal continues to be present and the time difference between the current recording duration and the target recording duration is less than the first duration, the system will continuously initiate new 5-second recordings. This approach ensures that the animal's complete behavior is captured in the initial recording, while also saving power by shortening the recording duration in subsequent recordings.

[0082] Optionally, the second preset duration can be gradually reduced based on the number of extensions, for example, the first extension is 5 seconds and the second extension is 4 seconds.

[0083] Through the above technical solution, this application achieves dynamic adjustment of recording duration during a single detection signal triggering process. The initial trigger uses a longer recording duration to ensure complete recording of the animal's initial behavior. Subsequent recordings use shorter recording durations to reduce power consumption.

[0084] In this embodiment, the first preset duration is not less than three seconds and not more than ten seconds.

[0085] As one embodiment, the solution of this application is implemented as follows: The first preset duration can be set to 5 seconds. After the hunting camera detects the infrared signal trigger, basic recording begins, and the recording duration is 5 seconds. This duration setting can cover the complete process of most animals' brief stays or behavioral changes, such as the animal entering the frame, briefly staying to observe the environment, and starting to forage. At the same time, the 5-second recording duration also avoids the waste of power caused by excessively long recordings, especially for scenarios where the animal has left or there are no target animals.

[0086] In one embodiment, such as Figure 3 As shown, step S200 also includes steps S230 to S250.

[0087] In this embodiment, step S230 involves inputting the acquired video footage into a preset recognition model.

[0088] In this embodiment, step S240 involves detecting animal images in the video footage using a preset recognition model and outputting the detection results.

[0089] In this embodiment, step S250 involves determining the object recognition result signal based on the detection result.

[0090] The preset recognition model can employ a convolutional neural network architecture. This model learns animal features, such as animal outlines, textures, and movement patterns, through a pre-trained dataset. The model's output detection results include the animal category, and a signal is triggered when a target animal category is detected. This preset recognition model is trained at the factory and stored in the corresponding control device for application to the hunting camera.

[0091] Specifically, basic recording begins after a passive infrared sensor is triggered, simultaneously capturing video footage at fixed intervals. The captured footage is transmitted in real-time to a pre-defined recognition model, which extracts image features using multi-layer convolutional kernels and compares them against a pre-stored target animal feature database. When a target animal category is detected and the confidence level is met, a target presence signal is generated, triggering the next basic recording session; if no target animal is detected or the confidence level is insufficient, a stop signal is generated to terminate recording.

[0092] The model's ability to analyze complex backgrounds, such as distinguishing between the movement of swaying tree branches and the trajectory of animal movements, effectively filters out non-target interference. The detection results are coordinated with a preset duration, for example, completing at least two image analyses within a 3-second basic recording cycle to ensure real-time judgment.

[0093] Through the above technical solution, this application achieves accurate identification and classification of animals in video footage. The preset recognition model provides a standardized recognition framework, avoiding the subjectivity of human judgment and the limitations of simple algorithms. The model's deep learning capability can effectively extract complex image features, improving the accuracy of animal detection.

[0094] In some of the solutions described above in this application, animal image detection is performed on the video footage by a preset recognition model and the detection results are output to determine the object recognition result signal. However, relying solely on the detection results of animal categories may lead to misjudgment. For example, when the detection model has low confidence in the recognition of the target animal category, it may still be judged as having a target animal, or non-target animal categories may be incorrectly classified due to detection errors, resulting in insufficient accuracy of the video control logic.

[0095] In one embodiment, such as Figure 4 As shown, step S250 further includes steps S251 and S253. The detection results include the animal category and the corresponding confidence level.

[0096] In this embodiment, step S251, if the animal category in the detection result is determined to be the target animal category and the corresponding confidence level is greater than a preset value, then the target animal category is determined to exist in the video footage.

[0097] In this embodiment, step S252, if the animal category in the detection result is the target animal category and the corresponding confidence level is not greater than a preset value, then it is determined that the target animal category does not exist in the video footage.

[0098] In this embodiment, step S253 is to determine that the animal category in the detection result is not the target animal category, and to determine that the target animal category does not exist in the video footage.

[0099] It should be noted that the recognition result signal includes two results: one is that the target animal category does not exist in the picture, and the other is that the target animal category exists in the picture. By step S251 to 253, it is determined whether the target animal category exists in the picture, and the corresponding recognition result signal is obtained.

[0100] The detection results are configured to include both the animal category and its corresponding confidence level parameter, with a preset value set as a threshold to distinguish between valid and invalid detections. For example, the preset value can be any value within the range of 60%-90%. When a target animal category is identified, its confidence level needs to be further verified to see if it exceeds the preset value; if not, it is considered an invalid detection. For non-target animal categories, they are directly excluded regardless of their confidence level. This dual-judgment mechanism, combined with the pre-set recognition model, can adapt to detection needs in different environments by dynamically adjusting the confidence threshold.

[0101] Specifically, after the preset recognition model outputs the detection results, animal category information is first extracted. If the category belongs to the preset target animal list, such as deer or bear, the confidence level is further compared with the preset value. When the confidence level is higher than the preset value, the presence of a target animal is determined and recording continues; if the confidence level is lower than or equal to the preset value, it is considered a false detection and recording stops. For non-target animal categories, such as birds or small rodents, even if the confidence level reaches 90% or higher, it is still determined that no target animal exists. By combining category filtering with confidence level verification, low-quality detection results caused by model misjudgment or environmental interference can be effectively filtered out, thereby reducing the generation of invalid recordings. For example, if the model misjudges a tree branch as a deer due to swaying branches, and the confidence level is only 50%, subsequent recording will not be triggered. This scheme significantly reduces the false trigger rate while maintaining detection sensitivity through dual condition constraints, thereby improving storage space utilization and device battery life.

[0102] Through the above technical solution, this application introduces a dual judgment mechanism of confidence threshold and animal category, optimizing the target animal identification logic. By filtering the confidence level using preset values, false detections with low confidence levels can be eliminated, avoiding invalid recordings due to model misjudgment. Furthermore, it improves the accuracy of target identification, reduces the risk of misoperation caused by model fluctuations or environmental interference, thereby effectively solving the problem of invalid data redundancy and improving the accuracy of recording control logic.

[0103] In one embodiment, such as Figure 5 As shown, this adaptive recording method also includes steps S610 and S620.

[0104] In this embodiment, step S610 involves continuously acquiring the video footage captured during the recording process while the hunting camera is in recording mode, and identifying the positions of objects in the acquired video footage to obtain a position recognition result signal.

[0105] In this embodiment, step S620 involves stopping recording after the current basic recording ends if the object in the video frame is determined to be farther from the hunting camera than a preset distance based on the location recognition result signal.

[0106] The location recognition result signal is generated through pixel ratio analysis of the target object. For example, if the pixel ratio of the target in the image is less than 1% of the total image area, it is determined to be too far away. The ranging module can be implemented using binocular vision ranging or a TOF sensor, and the preset distance can be set to an adjustable parameter within the range of 5 to 15 meters. The distance determination logic is time-series correlated with the basic recording cycle, ensuring that the stop command is executed only after a single basic recording is completed, avoiding video file fragmentation. The ranging data can be cross-validated with the target recognition result. When the target category is not a preset species, the distance determination step is skipped directly, reducing computational resource consumption.

[0107] Specifically, during continuous recording, target localization analysis is performed simultaneously. By calculating the spatial coordinates or relative size changes of the target object in real time, dynamic distance monitoring data is generated. When the target moves outside the preset monitoring range, the system maintains the integrity of the current basic recording until it ends naturally, and then cuts off the subsequent recording triggering link. This mechanism, while preserving valid behavioral segments, filters out long-distance false triggering scenarios through spatial dimensions. For example, when monitoring large animals, the preset distance can be set to 10 meters. When the target exceeds this range, the system terminates recording after completing the current 3-second basic recording.

[0108] In one feasible implementation, while the hunting camera is recording, it continuously acquires video footage captured during the recording process. The frame capture period can be set to 0.5 seconds, acquiring one frame every 0.5 seconds. After acquiring the video footage, a pre-trained deep learning model is used to identify the target animal type in the footage, obtaining a position recognition result signal. The position recognition result signal includes the object's coordinates and size information within the frame. Further, the distance of the target animal type relative to the hunting camera is calculated based on the position recognition result signal. Assuming a preset distance of 20 meters, if the calculated distance to the target animal type is greater than 20 meters, it is determined that the target animal type is farther from the hunting camera than the preset distance. Therefore, when it is determined that the object in the video footage is farther from the hunting camera than the preset distance, the hunting camera will stop recording after the current basic recording session ends.

[0109] Through the above technical solution, this application can effectively filter out invalid recordings triggered by distant targets, avoiding the generation of a large amount of useless data that occupies storage space. Furthermore, this solution can reduce energy consumption and extend the battery life of the hunting camera. Since continuous recording is only triggered by targets within the effective observation range, unnecessary recordings caused by distant targets are greatly reduced, thereby improving the quality and usability of the recorded data.

[0110] In some of the solutions described above in this application, the adaptive mechanism for controlling the recording duration based on the target animal identification result may cause the recording to continue indefinitely, resulting in the device being in a high power consumption state for a long time, which may cause excessive battery consumption, especially in scenarios where the animal is continuously active.

[0111] In one embodiment, such as Figure 6 As shown, this adaptive recording method also includes steps S710 and S720.

[0112] In this embodiment, step S710 is to start timing when the hunting camera starts recording.

[0113] In this embodiment, step S720 involves controlling the hunting camera to stop recording when the timer reaches the third preset duration.

[0114] The timing function is triggered synchronously with the recording operation, and the time accumulation is calculated through a hardware timing module or a software timer. The third preset duration can be configured to be 1 minute or 2 minutes, and the specific setting is not limited here. The third preset duration is pre-written by the device firmware or adjusted by the user interface.

[0115] The timing module continuously monitors the current cumulative recording time. When the preset threshold is reached, it sends an interrupt signal to the main control chip, triggering a command to forcibly stop recording.

[0116] This technical solution, together with the animal recognition-based video recording control logic in the preceding claim, forms a dual constraint mechanism: in scenarios where the animal is continuously present, continuous recording is terminated by a physical time limit; in scenarios where the animal appears intermittently, the original recognition control logic is retained. The combined use of the two maintains the continuity of behavior recording while avoiding the extreme power consumption that might result from a single control mechanism.

[0117] Specifically, when the passive infrared sensor triggers a detection signal to start basic recording, the timing module simultaneously begins accumulating the total recording time. Ideally, under conditions where the target animal is continuously detected, the device will cycle through the basic recording operation. During this time, the timing module continuously tracks the total recording time. When the accumulated value reaches the third preset duration, regardless of whether it is currently within a basic recording cycle, all recording processes are immediately terminated and the timer is reset. For example, in a scenario where the animal is continuously active for 2 minutes, if the third preset duration is set to 1 minute, the device will automatically shut down after recording the first minute of data to prevent the battery from being completely depleted.

[0118] Through the above technical solution, this application effectively solves the problem of excessive power consumption of the device in continuous detection scenarios. By introducing a physical timing forced termination mechanism, a hard upper limit on the total recording time is established while retaining the target animal recognition and control logic, thus avoiding the problem of infinite recording caused by repeated animal triggering.

[0119] This application also provides a control device that can implement the adaptive recording method described in the above embodiments. Therefore, the control device can at least achieve all the beneficial effects brought by the adaptive recording method in the above embodiments, which will not be elaborated here.

[0120] This application also provides a hunting camera that includes the control device described in the above embodiments. Therefore, the hunting camera can at least achieve all the beneficial effects brought by the control device described in the above embodiments, which will not be described in detail here.

[0121] Optionally, the control device may be integrated into the hunting camera and configured to implement the adaptive recording method described in the above embodiments.

[0122] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and not to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and they should all be covered within the scope of the claims and specification of this application. In particular, as long as there is no technical conflict, the various technical features mentioned in the various embodiments can be combined in any way. This application is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. An adaptive recording method, characterized in that, The adaptive recording method is applied to a hunting camera, which includes a passive infrared sensor. The adaptive recording method includes: When the passive infrared sensor triggers a detection signal, the hunting camera is controlled to start basic recording, which has an initial target recording duration; When the hunting camera is in recording mode, it continuously acquires the video footage captured during the recording process and identifies the objects in the acquired video footage to obtain the object recognition result signal. If it is determined, based on the object recognition result signal, that there is no target animal category in the video footage, the recording will stop after the current basic recording ends; If, based on the object recognition result signal, it is determined that a target animal category exists in the video footage, and the time difference between the current recording time of the hunting camera and the target recording time is less than a first time difference, the target recording time is extended.

2. The adaptive recording method as described in claim 1, characterized in that, The continuous acquisition of video footage captured during the basic recording process of the hunting camera while it is in recording mode includes: Obtain the image capture cycle; When the hunting camera is in recording mode, the video footage captured during the basic recording process of the hunting camera is obtained according to the aforementioned image capture cycle.

3. The adaptive recording method as described in claim 1, characterized in that, The adaptive recording method further includes: Set the initial target recording duration to the first preset duration; The step of extending the target recording time when it is determined that a target animal category exists in the video footage based on the object recognition result signal, and the time difference between the current recording time of the hunting camera and the target recording time is less than a first time difference, includes: If, based on the object recognition result signal, it is determined that there is a target animal category in the video footage, and the time difference between the current recording time of the hunting camera and the target recording time is less than the first time difference, the recording time is extended by a second preset time based on the target recording time, wherein the first preset time is greater than the second preset time.

4. The adaptive recording method as described in claim 3, characterized in that, The first preset duration is not less than three seconds and not more than ten seconds.

5. The adaptive recording method as described in claim 1, characterized in that, The process of obtaining the object recognition result signal from the acquired video footage includes: The acquired video footage is input into the preset recognition model; The video footage is used to detect animal images using the preset recognition model, and the detection results are output. The object recognition result signal is determined based on the detection results.

6. The adaptive recording method as described in claim 5, characterized in that, The detection results include the animal category and the corresponding confidence level; the determination of the object recognition result signal based on the detection results includes: If the animal category in the detection result is determined to be the target animal category, and the corresponding confidence level is greater than a preset value, then the target animal category is determined to exist in the video footage. If the animal category in the detection result is determined to be the target animal category, and the corresponding confidence level is not greater than a preset value, then it is determined that the target animal category does not exist in the video footage. If the animal category in the detection result is not the target animal category, it is determined that the target animal category does not exist in the video footage.

7. The adaptive recording method as described in any one of claims 1 to 6, characterized in that, The adaptive recording method further includes: When the hunting camera is in recording mode, it continuously acquires the video footage captured during the recording process and identifies the positions of objects in the acquired video footage to obtain the position recognition result signal. If, based on the location recognition result signal, it is determined that the object in the video footage is more than a preset distance from the hunting camera, recording will stop after the current basic recording ends.

8. The adaptive recording method as described in any one of claims 1 to 6, characterized in that, The adaptive recording method further includes: The timer starts when the hunting camera begins recording; Once the timer reaches the third preset duration, control the hunting camera to stop recording.

9. A control device, characterized in that, The control device is used to implement the adaptive recording method as described in any one of claims 1 to 8.

10. A hunting camera, characterized in that, The hunting camera includes the control device as described in claim 9.

Images