Image stabilization for a wearable camera device

Abstract

To provide a wearable camera device having an image stabilization function in relation to a camera device using image stabilization.SOLUTION: An image capture unit (101) for capturing a sequence of images, comprising at least a lens (103) and an image sensor (102), and comprising a vibration sensor (104) for measuring vibrations, an optical image stabilization mechanism (105) for physically compensating for vibrations by physically moving the lens and / or the image sensor, an electronic image stabilization circuit (106) for digitally compensating for vibrations by image processing, and a processing circuit configured to select between the optical image stabilization mechanism and the electronic image stabilization circuit by comparing camera vibrations with a predetermined stabilization margin and / or based on a position of an object in the sequence of images.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a camera device with image stabilization. The present disclosure also relates to a method for controlling a wearable camera device with image stabilization functionality. [Background technology]

[0002] Electronic image stabilization (EIS) and optical image stabilization (OIS) are the two primary technologies used in modern wearable camera devices to reduce the effects of camera vibration and movement, thereby improving image quality. EIS relies on an onboard sensor, typically a gyroscope sensor, and associated image transformation acceleration functions on the processing chip to digitally compensate for camera vibration. This method is advantageous due to its low cost and reduced power consumption, as it does not require additional mechanical components. However, EIS has inherent limitations, such as a limited stabilization margin.

[0003] On the other hand, OIS uses mechanical components to physically move the lens and / or image sensor in response to detected camera vibration. This method allows for more significant stabilization. However, the inclusion of mechanical components makes OIS more expensive and power-intensive compared to EIS. Furthermore, the mechanical components in OIS can wear out more quickly, especially when operated continuously over long periods of time. This mechanical wear and tear poses a significant drawback, especially in applications requiring long periods or intensive use.

[0004] Given these limitations, current body-worn camera devices must balance the cost and power efficiency of EIS with the superior stabilization capabilities of OIS. Neither technology alone can provide an optimal solution for all scenarios, especially in environments with varying degrees of camera vibration. Users are often forced to compromise between image stability and the longevity and cost-effectiveness of their camera equipment. This compromise is particularly evident in professional settings, where high-quality image stabilization is important, but operational costs and equipment durability must also be considered.

[0005] It is therefore an object of the present disclosure to provide an improved body-worn camera device that integrates both electronic and optical image stabilization techniques. How to best combine electronic and optical image stabilization techniques remains a challenge. Summary of the Invention

[0006] The present disclosure provides: an image capture unit for capturing a sequence of images, the image capture unit comprising at least a lens and an image sensor; a vibration sensor for measuring camera vibration; an optical image stabilization mechanism for physically compensating for camera shake by physically moving the lens and / or image sensor; an electronic image stabilization circuit for digitally compensating for camera shake by image processing; a processing circuit configured to select between the optical image stabilization mechanism and the electronic image stabilization circuit by comparing the camera vibration to a predetermined stabilization margin and / or based on the position of the object within the sequence of images; The present invention relates to a wearable camera device comprising:

[0007] The combination of optical and electronic image stabilization in the proposed camera device offers significant technical advantages. By integrating both stabilization mechanisms, the device can dynamically select the most effective method based on real-time measurements of camera vibration. This hybrid approach allows the camera's stabilization performance to be optimized according to the specific conditions encountered by the camera, ensuring high-quality image capture across a wide range of environments. A vibration sensor measures camera vibration. For example, if camera vibration is within the manageable range of electronic image stabilization, the device can prioritize this method to save power and reduce mechanical wear. Conversely, in situations where vibration exceeds the capabilities of electronic stabilization, the device can switch to optical stabilization for more effective compensation.

[0008] The processing circuit's ability to compare measured camera vibrations with a predetermined stabilization margin or to determine the position of an object within a sequence of images improves the adaptability and accuracy of the device, while at the same time extending the camera's operational life by minimizing its reliance on mechanical components. This capability ensures that the most appropriate stabilization method is selected based on the actual dynamics of the scene, providing more tailored and effective stabilization.

[0009] The proposed method for controlling a body-worn camera device involves acquiring a sequence of images, measuring camera vibrations, and choosing between compensating for these vibrations by physically moving the lens and / or image sensor or by digitally processing the image. The choice is based on comparing the measured vibrations to a predetermined stabilization margin and / or the position of the object in the image. This method ensures that the camera can effectively adapt to various conditions and provides stable, high-quality video capture.

[0010] The vibration sensor can include a gyroscope and / or accelerometer capable of measuring angular motion between successive images. The electronic image stabilization circuit can compensate for vibration by shifting rows or columns in the image, and the predetermined stabilization margin can correspond to a cropping margin or empty margin in the image for digital compensation. The processing circuit can select optical stabilization for vibrations above a predetermined margin, electronic stabilization for vibrations within a predetermined margin, or a combination of both methods for optimal performance.

[0011] The device's processing circuitry can also identify and determine the location of the object, which influences the selection of the stabilization method. For example, when using electronic stabilization, if the object is within a predetermined area that is not affected by added free space, the device may prioritize digital compensation. Alternatively, if the object overlaps the stabilization margin, the device may select optical stabilization.

[0012] The present disclosure relates to a method for controlling a body-worn camera device having optical and electronic image stabilization capabilities, the method comprising: acquiring a sequence of video images from a body-worn camera device; measuring camera vibrations of a body-worn camera device; by comparing the camera vibration with a predetermined stabilization margin and / or based on the position of the object in the sequence of images; Compensating for camera vibration by physically moving the lens and / or image sensor of the wearable camera device; and Digitally compensate for camera vibration through image processing To choose either Includes.

[0013] Those skilled in the art will recognize that the disclosed methods for controlling a body-worn camera device having optical and electronic image stabilization capabilities can be performed using any embodiment of the disclosed body-worn camera device, and vice versa.

[0014] Various embodiments are described below with reference to the drawings, which are exemplary embodiments and intended to illustrate some of the features of the disclosed body-worn camera devices and methods for controlling body-worn camera devices, but are not intended to limit the invention. [Brief explanation of the drawings]

[0015] [Figure 1] FIG. 1 illustrates one embodiment of a wearable camera device of the present disclosure. [Figure 2] FIG. 1 is a diagram of an image with a stabilization margin and possible fields of view. [Figure 3] FIG. 1 shows an example of an image including an object within a predetermined analysis area. [Figure 4A] FIG. 1 is a diagram showing an example of an image including an object and surrounding cross lines. [Figure 4B] FIG. 1 is a diagram showing an example of an image including an object and surrounding cross lines. [Figure 5] FIG. 10 shows an example of the empty space added to some lines of an image when using electronic image stabilization. [Figure 6] 1 is a flowchart of a method according to one embodiment of the disclosed method for controlling a body-worn camera device having optical and electronic image stabilization capabilities. DETAILED DESCRIPTION OF THE INVENTION

[0016] The present disclosure provides: an image capture unit for capturing a sequence of images, the image capture unit comprising at least a lens and an image sensor; a vibration sensor for measuring camera vibration; an optical image stabilization mechanism for physically compensating for camera shake by physically moving the lens and / or image sensor; an electronic image stabilization circuit for digitally compensating for camera shake by image processing; a processing circuit configured to select between the optical image stabilization mechanism and the electronic image stabilization circuit by comparing the camera vibration to a predetermined stabilization margin and / or based on the position of the object within the sequence of images; The present invention relates to a wearable camera device comprising:

[0017] 1 illustrates one embodiment of a wearable camera device 100 of the present disclosure, comprising an image capture unit 101 that includes a lens 103 and an image sensor 102. The wearable camera device 100 of FIG. 1 further comprises a vibration sensor 104, an optical image stabilization mechanism 105, an electronic image stabilization circuit 106, and a processing circuit 107.

[0018] The body worn camera devices described in this disclosure can be implemented in a variety of forms to suit different applications and environments.

[0019] As an example, the wearable camera device described in the present disclosure includes a fixed structure and can be mounted to various surfaces and structures, including, for example, a bar, a stick, or a building. As an example, the camera device can be designed with a universal mounting system including an adjustable clamp or bracket. These components allow the device to be securely attached to a cylindrical object such as a bar or stick, making it ideal for use in action sports, vlogs, or other dynamic environments where handheld or pole-mounted cameras are commonly used. Generally, those skilled in the art know how to attach or mount a camera, such as a surveillance camera, to a fixed structure.

[0020] For mounting to a building or other fixed structure, the camera device may feature a robust mounting plate or screw-based fasteners. These can be used to securely attach the device to a wall, ceiling, or other surface, providing stable and reliable performance in surveillance, construction site monitoring, or outdoor security applications. The camera device's housing in this scenario may be weather-resistant and dust-proof, ensuring it can withstand harsh environmental conditions over long periods of time.

[0021] Additionally, magnetic mounts can be an option for temporary installation on metal surfaces, offering quick and easy installation and removal. This type of mounting system is particularly useful for industrial inspection, maintenance inspections, or any application where the camera needs to be frequently repositioned or adjusted.

[0022] In one embodiment, the body-worn camera device may be integrated into a more rugged housing designed for industrial or security applications. This variation may include additional features such as weather resistance, dust resistance, and a reinforced casing to ensure reliable operation in harsh conditions. Such a design would be beneficial for surveillance systems, outdoor monitoring, and other scenarios where environmental exposure is a concern.

[0023] Additionally, the processing circuitry within the camera device may be designed to support a variety of computing capabilities and connectivity options, including, but not limited to, Ethernet and Power over Ethernet (PoE) connections and communications equipment. In one embodiment, the processing unit includes advanced image processing algorithms and machine learning capabilities to improve stabilization performance and automatically adapt to different shooting conditions. Connectivity options may also include wireless communication protocols, such as Wi-Fi, Bluetooth, or cellular networks, enabling remote control, real-time video streaming, and seamless integration with other devices or systems. Optical Image Stabilization (OIS)

[0024] Optical image stabilization mechanisms in body-worn camera devices are designed to physically compensate for camera vibration by adjusting the position of the lens and / or image sensor. This technology can reduce motion blur and maintain image clarity, especially in environments with significant or unpredictable motion.

[0025] The term OIS within the context of this disclosure shall be broadly interpreted to include adjustment of the lens or image sensor, or a combination thereof. Adjusting the position of the image sensor is commonly referred to as sensor image stabilization (SIS), which is considered a variant of OIS in this context. This technique is otherwise very similar to adjusting the position of the lens.

[0026] In one embodiment, the OIS mechanism includes a set of gyroscopes and actuators. The gyroscope detects the angular movement of the camera and provides real-time data about that movement. This information is then processed by the camera's processing circuitry, which calculates the necessary adjustments to counteract the detected vibrations. The actuators, which may be based on piezoelectric or electromagnetic principles, move the lens or sensor accordingly, ensuring that the image remains stable despite camera movement.

[0027] OIS mechanisms can be designed with various degrees of freedom. Typical use cases require at least two axes to compensate for vertical and horizontal (pitch and yaw) motion. On the other hand, multi-axis OIS, with at least three axes, can address motion around three axes, including pitch, yaw, and roll, providing more comprehensive stabilization.

[0028] In another implementation, the OIS mechanism can involve lens-shift technology, in which a lens element is moved to compensate for camera shake. Alternatively, sensor-shift technology can be used, in which the image sensor itself is moved to counteract the vibrations. Some advanced systems may even combine both lens-shift and sensor-shift methods to maximize the stabilization effect.

[0029] The accuracy of the OIS mechanism is enhanced by advanced control algorithms that predict and react to camera movement with high precision. These algorithms can be based on a combination of gyroscope data and other sensor inputs, such as accelerometers, to provide a more comprehensive understanding of camera movement. This predictive capability ensures that the OIS mechanism can quickly and effectively counter vibrations, even in high-frequency or high-amplitude scenarios. Electronic Image Stabilization (EIS)

[0030] Electronic image stabilization circuits in wearable camera devices are designed to digitally compensate for camera shake through image processing. This approach improves the stability of captured images and videos by correcting unwanted motion artifacts in real time without the need for mechanical adjustments.

[0031] An EIS system may rely on one or more gyroscopes and / or one or more accelerometers. In one embodiment, an EIS system relies on a combination of sensors, such as gyroscopes and accelerometers, to measure camera motion. These sensors provide accurate data regarding the direction and magnitude of vibrations. The data is then processed by the camera's processing circuitry, which calculates the necessary adjustments to the captured image to counteract the detected motion.

[0032] The electronic image stabilization circuit for performing the electronic image stabilization may be any processing circuit suitable for the task, for example, it may be a separate unit or part of a general processing circuit.

[0033] EIS systems typically include image processing algorithms that digitally stabilize the image. One common technique involves shifting the image frame based on measured motion data. For example, if the camera moves to the right, the EIS system shifts the image frame to the left by a corresponding amount, possibly line by line, thereby nullifying the effect of the motion. This process is typically performed in real time, ensuring that each frame is adjusted before being displayed or recorded.

[0034] In another implementation, the EIS circuitry can use frame interpolation methods to smooth transitions between frames. This technique involves analyzing multiple frames to detect and predict motion patterns, allowing the system to create intermediate frames that provide a smoother visual experience. Frame interpolation is particularly useful in scenarios involving rapid motion, as it helps eliminate the jitter effect often seen in such situations.

[0035] EIS systems may also use rolling shutter correction techniques. Rolling shutter artifacts occur when different parts of the image sensor capture a scene at slightly different times, resulting in distortions in fast-moving scenes. EIS circuitry can detect and correct these distortions by aligning captured frames based on the sensor's readout timing, resulting in a more stable and accurate representation of the scene.

[0036] EIS circuitry can be configured to work in conjunction with other image processing functions, such as noise reduction, exposure control, and color correction. By integrating these functions, the system ensures that the stabilization process does not adversely affect overall image quality. For example, an EIS system can be designed to maintain, to some extent, the sharpness and clarity of the stabilized image, even when significant adjustments are made to compensate for vibration.

[0037] In one embodiment, the EIS system includes machine learning algorithms that adapt to different shooting conditions and improve over time. These algorithms analyze patterns of camera movement and the resulting image data to optimize the stabilization process. By continually learning from new data, the EIS system can improve its performance and provide more accurate stabilization in a wider range of scenarios.

[0038] The wearable camera device described in this disclosure includes an electronic image stabilization (EIS) circuit designed to compensate for camera vibration by digitally processing captured images. In one embodiment, vibration is compensated for by shifting one or more rows or columns in at least one image of a sequence of images. For example, an image can be thought of as several rows. Electronic image stabilization can then render the image one row at a time. For each row, horizontal compensation for vibration can be performed. This can be seen as empty space or black space. Figure 5 shows an example of empty space 115 added to some lines of an image 108 when using electronic image stabilization.

[0039] In one embodiment, the EIS circuitry detects camera vibration using data from vibration sensors, such as a gyroscope and an accelerometer. Based on this data, the EIS system determines the direction and magnitude of camera movement. Once the movement is quantified, the EIS circuitry adjusts the position of captured image frames to counteract this movement.

[0040] The adjustment process can involve shifting the image content horizontally or vertically by moving rows or columns within the image. For example, if the camera shakes slightly to the left, the EIS system compensates by shifting the image to the right. This shift realigns the image content with the original intended frame, thereby reducing the effects of the vibration.

[0041] Shifting rows or columns in an image can be achieved by real-time image processing algorithms. These algorithms analyze successive frames to detect misalignments caused by vibrations and apply the necessary shifts to correct them. This process is typically performed frame-by-frame, ensuring that each image in the sequence is stabilized before being displayed or recorded. Vibration Sensor

[0042] The body-worn camera devices described in this disclosure include one or more vibration sensors, which may include one or more gyroscopes and / or one or more accelerometers, that provide the data necessary for both optical image stabilization (OIS) and electronic image stabilization (EIS) systems.

[0043] The one or more vibration sensors may be configured to measure angular motion of the body-worn camera between successive images of the sequence of images.

[0044] In one embodiment, the vibration sensor includes at least one gyroscope. The gyroscope measures angular velocity, which refers to the rate of rotation around the camera's axis. By detecting how fast and in which direction the camera is rotating, the gyroscope provides precise information about the camera's orientation and movement. This data is necessary for OIS systems because it allows actuators to precisely adjust the lens or image sensor to counteract rotational vibrations. In EIS systems, gyroscope data helps digitally align image frames to compensate for rotational motion.

[0045] Additionally, vibration sensors can include an accelerometer. The accelerometer measures linear acceleration, which refers to the rate of change of velocity along the camera's linear axes (x, y, and z directions). This sensor detects movements such as tilt, pan, and shake, providing comprehensive data on camera movement. Accelerometer data complements gyroscope data by capturing both rotational and translational movement, providing a more complete picture of camera behavior. For OIS, this means that both rotational and linear vibrations can be physically compensated for by moving the lens or image sensor accordingly. For EIS, the accelerometer helps provide precise digital correction for image stabilization.

[0046] Those skilled in the art will know how to implement suitable vibration sensors. For example, gyroscopes and accelerometers can be implemented using microelectromechanical systems (MEMS) technology, which offers high accuracy in a compact form factor. MEMS sensors are well suited for body-worn camera devices due to their small size, low power consumption, and durability. Image stabilization options

[0047] A stabilization margin in the context of a wearable camera device can be thought of as an area or margin inside the original outer image frame. The concept is illustrated in Figure 2, where image 108 is an image from a sequence of images. Image 108 has a stabilization margin 109. Arrow 110 represents the amplitude of camera vibration, typically translated into horizontal movement of an entire row of pixels. An example of a field of view is shown as a dotted line in Figure 2. The inner line 111 can be seen as an analysis area that is not affected by the added free space of the image sequence when using electronic image stabilization circuits.

[0048] In one embodiment, the processing circuitry is configured to select to use optical image stabilization when the camera vibration has an amplitude greater than a predetermined stabilization margin, and to select to use electronic image stabilization or a combination of electronic image stabilization and optical image stabilization when the camera vibration has an amplitude less than the predetermined stabilization margin.

[0049] The stabilization margin can correspond to a cropping margin or empty margin in a sequence of images for digital compensation of camera vibration. When a camera captures an image, it can be seen that the image contains extra content around the edges beyond what was intended for the final frame. This extra content can be said to form the cropping margin. If the camera moves or shakes during the stabilization process, the EIS circuitry can adjust the position of the main image within this margin to counteract the movement. By shifting the image content horizontally or vertically within the cropping margin, the system keeps the final image stable and centered.

[0050] The predetermined stabilization margin may be defined as a wearable camera device rotation angle selected from the range of 1 to 30% of the camera field of view, for example, about 20% of the horizontal field of view.

[0051] The wearable camera device described in this disclosure can include processing circuitry configured to select the most appropriate stabilization method based on the amplitude of detected camera vibration relative to a predetermined stabilization margin. This functionality enables the camera device to optimize image stabilization by dynamically selecting between optical image stabilization (OIS), electronic image stabilization (EIS), or a combination of both, depending on the intensity of the vibration.

[0052] In one embodiment, the processing circuitry continuously receives data from vibration sensors, such as a gyroscope and an accelerometer, to estimate the amplitude of camera movement. Amplitude refers to the magnitude of the vibration and can vary from small, subtle movements to larger, more noticeable shifts. Amplitude can be converted to distance on the image.

[0053] The processing circuitry may be configured to select the OIS mechanism if the detected vibration has an amplitude greater than a predetermined stabilization margin. The predetermined stabilization margin serves as a threshold level, and vibrations exceeding this threshold indicate that the movement is too large to be effectively handled by EIS alone. OIS, which involves physically moving the lens and / or image sensor, can better compensate for larger amplitude vibrations because it directly counteracts camera movement through mechanical adjustments. This method ensures that the image remains stable even during substantial movement, providing superior stabilization in high-vibration scenarios.

[0054] Conversely, if the detected vibration has an amplitude below a predetermined stabilization margin, the processing circuitry chooses to use the EIS circuitry or a combination of EIS and OIS. For small-amplitude vibrations, digital compensation with EIS may be sufficient. EIS adjusts the image by shifting rows or columns within the frame, effectively stabilizing the image without requiring mechanical movement. This method is advantageous for saving power and reducing mechanical wear on the OIS components, thus extending the overall lifespan of the camera device.

[0055] The processing circuitry can be configured to identify the object and determine the object's location, which can be done using a variety of image processing and computer vision techniques, as will be appreciated by those skilled in the art.

[0056] Image processing and computer vision techniques can identify objects using feature detection algorithms. Feature detection involves analyzing an image to find distinctive points or patterns, such as edges, corners, or textures, which can be used to recognize and track objects. To identify these features, algorithms such as the Harris Corner Detector, Scale Invariant Feature Transform (SIFT), or Speed-Up Robust Features (SURF) can be used. Once detected, the system can match them across successive frames to track the object's movement and determine its location within the image.

[0057] Another approach involves using object recognition techniques that rely on machine learning models trained to recognize specific objects. These models can be based on convolutional neural networks (CNNs), which are particularly effective for image classification and object detection tasks. The processing circuitry can be equipped with pre-trained CNN models such as YOLO (You Only Look Once), SSD (Single Shot Multibox Detector), or Faster R-CNN, which can accurately identify and localize objects within a frame.

[0058] In some implementations, the processing circuitry can identify objects using a motion detection algorithm. This method involves analyzing the difference between successive frames to detect moving objects. Techniques such as background subtraction, optical flow, or frame differencing can be used to identify areas of the image that may contain motion. These areas are then analyzed to determine whether they correspond to objects. This approach is particularly useful in dynamic environments where objects are typically moving, such as sports or surveillance applications.

[0059] The processing circuitry can also use advanced tracking algorithms to maintain a consistent lock on an object once it has been identified. Methods such as Kalman filters, particle filters, or correlation-based trackers can predict the object's motion and update its position in real time.

[0060] More specifically, the processing circuitry may be configured to determine the position of the object relative to a stabilization margin, or a predetermined region, or a surrounding cross line.

[0061] In one embodiment, when using EIS to digitally compensate for camera vibration, the processing circuitry is configured to select EIS if the object is within a predetermined analysis area positioned to avoid the impact of additional empty space in the image sequence. This predetermined analysis area may be, for example, an area where all or certain individuals cannot linger without valid reason. In such applications, the processing circuitry can measure the time the object remains within the area. If the time exceeds a predetermined time threshold, a loitering alert can be generated. By selecting EIS in these situations, the system can take advantage of the benefits of digital stabilization, such as reduced mechanical wear and power consumption, while ensuring that the object remains clear and stable within the image frame.

[0062] Additionally, the processing circuitry can be configured to select EIS when the object is within a region within the stabilization margin. As previously mentioned, the stabilization margin is a buffer zone around the edge of a captured image used for digital compensation. When an object is located within this margin, it indicates that there is enough space around the object for the EIS system to make necessary adjustments without compromising the object's visibility or clarity. In such cases, the processing circuitry can prioritize EIS to maintain stabilization efficiency and extend the life of OIS components.

[0063] In the active mode, the predetermined stabilization margin can be defined as a rotation angle of the wearable camera device selected from the range of 1 to 30% of the field of view. When the wearable camera device is in the passive mode, the predetermined stabilization margin can be set to a larger value than in the active mode.

[0064] Conversely, the processing circuitry may be configured to select the OIS mechanism when the object overlaps a predetermined stabilization margin. If the object is near or within the stabilization margin, digital adjustments are more likely to push parts of the object out of the frame or affect its clarity. In these scenarios, using OIS, which involves physically moving the lens or image sensor, ensures that the entire scene, including the object, remains stable within the frame.

[0065] The body-worn camera devices described in this disclosure may be implemented with configurable modes of operation. Specifically, the devices can operate in active or passive modes, each tailored to different use cases and operational needs.

[0066] In one embodiment, the body-worn camera device can be configured into an active mode, in which the device transmits a video stream in real time. This mode is particularly useful in surveillance systems where immediate visual feedback is useful. For example, if a security guard, operator, or other personnel is alerted to an event or suspicious activity in a monitored area, the camera can be switched into active mode. In this mode, a live video feed is sent directly to the operator, enabling real-time monitoring and rapid decision-making. Active mode ensures that the operator receives the most up-to-date visual information, allowing the operator to respond quickly to incidents.

[0067] The active mode can utilize both optical image stabilization (OIS) and electronic image stabilization (EIS) systems to provide a stable and clear video stream. The processing circuitry can dynamically select the most appropriate stabilization method based on detected camera vibration and the position of an object in the frame. For example, when the wearable camera device is in the active mode, the processing circuitry may be configured to select to use the optical image stabilization mechanism when the camera vibration has an amplitude greater than a predetermined stabilization margin, and to select to use the electronic image stabilization circuit when the camera vibration has an amplitude less than the predetermined stabilization margin.

[0068] Furthermore, when the wearable camera device is in active mode, the processing circuitry may be configured to select to use optical image stabilization if the wearable camera device operates in high dynamic range (HDR) mode. HDR mode improves image quality by capturing and combining multiple exposures to create an image with a wider range of brightness levels. HDR mode is particularly beneficial in scenes with both very bright and very dark areas, as it helps preserve detail in both highlights and shadows, resulting in a more balanced and visually appealing image.

[0069] In another embodiment, the body-worn camera device can be configured in a passive mode, in which the device primarily analyzes video data rather than transmitting it in real time. Passive mode is suitable for applications where continuous video streaming is not required and focuses on processing and analyzing recorded footage. This mode can be used for post-event analysis, long-term monitoring, or automated surveillance tasks where real-time human intervention is not required.

[0070] In passive mode, the camera device can use advanced video analytics to detect and track objects, identify patterns, and recognize events or anomalies in the captured footage. The processing circuitry can efficiently analyze the video data using machine learning algorithms and other image processing techniques.

[0071] When the camera device is in passive mode, the processing circuitry can be configured to select EIS if the object is within a predetermined analysis area that is affected by additional empty space in the image sequence. EIS involves digitally compensating for camera vibration by shifting rows or columns within the image frame. This digital adjustment can sometimes result in empty space or margins around the edges of the image. EIS can be used for analysis areas that do not intersect with these empty spaces.

[0072] In one embodiment, the processing circuitry is configured to select the EIS when the camera device is set to analyze the video data only within a predetermined analysis area that is not affected by empty spaces created during the EIS digital compensation process. This configuration may be useful for tasks that involve tracking or analyzing specific areas of a frame, such as monitoring a specific area in a security system or focusing on a specific zone in an industrial inspection process.

[0073] 3 shows an example of an image 108 that includes an object 113 within a predetermined analysis area 112. As can be seen, the predetermined analysis area 112 does not overlap with the stabilization margin 109, allowing EIS to be used.

[0074] The wearable camera devices described in this disclosure may include processing circuitry that enhances their image stabilization capabilities by intelligently responding to the movement of objects within a captured scene. One such feature involves detecting an object crossing a perimeter crossing line and corresponding adjustment of the stabilization method based on this movement.

[0075] In one embodiment, the processing circuitry is configured to detect whether an object crosses a perimeter crossing line, which is a virtual boundary defined within the camera's field of view and can be set by a user or pre-configured based on application requirements. The crossing line can be used to monitor specific areas within a scene, such as entry points in a security surveillance setting or critical areas in an industrial monitoring scenario.

[0076] When an object crosses a predetermined crossing line, the system can register this event and trigger a specific response. This functionality is particularly useful for alerting operators to movement in critical areas or automating responses to specific movements detected within a monitored environment.

[0077] Additionally, the processing circuitry can be configured to select an optical image stabilization (OIS) mechanism when the surrounding cross lines overlap a predetermined stabilization margin. The predetermined stabilization margin, as previously described, is a buffer zone around the edge of the captured image used for digital compensation by an electronic image stabilization (EIS) system. If an object is detected near or within this margin, digital image stabilization adjustments risk pushing part of the object out of the frame or affecting its clarity.

[0078] By choosing OIS in such situations, the camera device ensures more precise stabilization through physical adjustments of the lens or image sensor, maintaining image integrity and keeping the subject fully visible and clear. OIS offers a mechanical approach to stabilization that does not rely on digital cropping or shifting, thus preserving the full field of view.

[0079] This selection process, based on the detection of crossing the perimeter crossing line, increases the reliability and effectiveness of the camera's stabilization system. For example, in security applications, when a person enters a restricted area defined by the perimeter crossing line, the camera can switch to OIS to ensure the best possible image quality for identifying the individual. Similarly, in industrial settings, when machinery or equipment crosses a critical operating area, the camera can use OIS to provide clear, stabilized footage for monitoring or analysis.

[0080] 4A and 4B show example images 108 that include an object 113 and surrounding cross lines 114. In FIG. 4A, the surrounding cross lines 114 do not overlap the stabilization margin 109. In FIG. 4B, there are two surrounding cross lines 114, both of which overlap the stabilization margin 109.

[0081] The wearable camera device may include processing circuitry for performing any of the disclosed processing steps. As will be appreciated by those skilled in the art, the processing circuitry may be any suitable processing circuitry, such as a single processor or a multi-core / multi-processor system, provided, for example, as part of a portable system or as a separate component. The system may further include peripheral components, such as one or more memories that may be used to store instructions that may be executed by any of the processors. The one or more memories may include random access memory (RAM) and / or read-only memory (ROM), or any suitable type of memory. The system may further include internal and external network interfaces, input and / or output ports, modules for wirelessly transmitting and receiving data, and / or communication interfaces that enable the transfer of software and / or data between the system and external devices.

[0082] The present disclosure further relates to a method for controlling a body-worn camera device having optical and electronic image stabilization capabilities.

[0083] 6 shows a flowchart of a method 200 according to one embodiment of the disclosed method for controlling a body-worn camera device having optical and electronic image stabilization capabilities. The method can include acquiring a sequence of video images from the body-worn camera device (201), measuring camera vibration of the body-worn camera device (202), and selecting between compensating for the camera vibration by physically moving the lens and / or image sensor of the body-worn camera device and digitally compensating for the camera vibration by image processing (203) by comparing the camera vibration to a predetermined stabilization margin and / or based on the position of an object in the sequence of images.

[0084] The present disclosure further relates to a computer program having instructions that, when executed by a computing device or computing system, cause the computing device or computing system to perform a method for controlling a body-worn camera device. The computer program may be stored on any suitable type of storage medium, such as a non-transitory storage medium.

Claims

1. an image capture unit (101) for capturing a sequence of images (108), the image capture unit (101) comprising at least a lens (103) and an image sensor (102); a vibration sensor (104) for measuring camera vibration; an optical image stabilization mechanism (105) for physically compensating for the camera shake by physically moving the lens (103) and / or the image sensor (102); an electronic image stabilization circuit (106) for digitally compensating for said camera shake by image processing; a processing circuit (107) configured to select between the optical image stabilization mechanism (105) and the electronic image stabilization circuit (106) by comparing the camera shake with a predetermined stabilization margin (109) and / or based on the position of an object (113) within the sequence of images (108); A wearable camera device (100) comprising:

2. 2. The wearable camera device (100) of claim 1, wherein the predetermined stabilization margin (109) corresponds to a cropping margin or empty margin in the sequence of images (108) for digital compensation of the camera vibration.

3. 3. The wearable camera device (100) of claim 1 or 2, wherein the processing circuit (107) is configured to select to use the optical image stabilization mechanism (105) when the camera vibration has an amplitude greater than the predetermined stabilization margin (109), and to select to use the electronic image stabilization circuit (106) or a combination of the electronic image stabilization circuit (106) and the optical image stabilization mechanism (105) when the camera vibration has an amplitude less than the predetermined stabilization margin (109).

4. 4. The wearable camera device (100) of claim 1, wherein the processing circuit (107) is configured to identify the object (113) and determine the position of the object (113).

5. 5. The wearable camera device (100) of claim 4, wherein the processing circuitry (107) is configured to determine the position of the object (113) relative to the stabilization margin (109) or a cross line (114) of a predetermined area or perimeter.

6. 6. The wearable camera device (100) of claim 1, wherein the processing circuit (107) is configured to select to use the electronic image stabilization circuit (106) when using the electronic image stabilization circuit (106) to digitally compensate for the camera shake if the object (113) is within a predetermined analysis area arranged so as not to be affected by additional empty space in the sequence of images (108).

7. 7. The wearable camera device (100) of claim 1, wherein the processing circuit (107) is configured to select to use the electronic image stabilization circuit (106) when the object (113) is within an area inside the stabilization margin (109).

8. 8. The wearable camera device (100) of claim 1, wherein the processing circuit (107) is configured to select to use the optical image stabilization mechanism (105) when the object (113) overlaps the predetermined stabilization margin (109).

9. 9. The wearable camera device (100) of claim 1, wherein the wearable camera device (100) is configurable in an active mode in which the wearable camera device (100) transmits a video stream and is further configurable in a passive mode in which the wearable camera device (100) analyzes video data.

10. 10. The wearable camera device (100) of claim 9, wherein when the wearable camera device (100) is in the active mode, the processing circuit (107) is configured to select to use the optical image stabilization mechanism (105) if the camera vibration has an amplitude greater than the predetermined stabilization margin (109), and to select to use the electronic image stabilization circuit (106) if the camera vibration has an amplitude less than the predetermined stabilization margin (109).

11. 11. The wearable camera device (100) of claim 9 or 10, wherein when the wearable camera device (100) is in the active mode, the processing circuit (107) is configured to select to use the optical image stabilization mechanism (105) if the wearable camera device (100) operates in a high dynamic range mode.

12. 12. The wearable camera device (100) of claim 9, wherein when the wearable camera device (100) is in the passive mode, the processing circuit (107) is configured to select to use the electronic image stabilization circuit (106) to digitally compensate for the camera shake if the object (113) is within a predetermined analysis area arranged so as not to be affected by additional empty space in the sequence of images (108).

13. 13. The wearable camera device (100) of claim 9, wherein when the wearable camera device (100) is in the active mode, the predetermined stabilization margin (109) is defined as a rotation angle of the wearable camera device (100) and is selected from a range of 1 to 30% of a field of view, and when the wearable camera device (100) is in the passive mode, the predetermined stabilization margin (109) is set to a value greater than in the active mode.

14. A method (200) for controlling a body-worn camera device (100) with optical and electronic image stabilization capabilities, comprising: acquiring (201) a sequence of video images (108) from the wearable camera device (100); measuring (202) camera vibrations of the wearable camera device (100); by comparing the camera shake with a predetermined stabilization margin (109) and / or based on the position of an object (113) within the sequence of images; - compensating for said camera vibration by physically moving the lens (103) and / or image sensor (102) of said wearable camera device (100); and - digitally compensating for said camera vibrations by image processing Selecting either (203) A method (200) comprising:

15. 15. A non-transitory storage medium containing a computer program having instructions that, when executed by a computing device or system, cause the computing device or system to perform the method for controlling a body-worn camera of claim 14.

Images