METHODS TO ACHIEVE SMOOTH ZOOM FOR VIDEO IN CAMERAS

    公开(公告)号:US20240080564A1

    公开(公告)日:2024-03-07

    申请号:US18173443

    申请日:2023-02-23

    Applicant: Apple Inc.

    CPC classification number: H04N23/69 H04N23/45 H04N23/6842

    Abstract: Electronic devices, methods, and program storage devices for achieving smooth zooming operations during video capture are disclosed. In particular, smooth zooming may be desirable during video capture operations that involve a single image capture device and/or transitioning between two or more distinct image capture devices, e.g., having different optical and/or zooming properties, during the video capture. When video zooming is done abruptly, it can lead to an unpleasant user experience. The techniques described herein to improve the smoothness of zooming operations include: the use of longer zoom “ramps” for image capture devices; the “early” transitioning between image capture devices during video capture; and the performance of additional “digital zoom smoothing-aware” video image stabilization operations. The embodiments described herein also provide for a more consistent user experience between video streaming (i.e., “preview”) modes and the recorded (i.e., “final”) zoomed video that is produced using the various zoom smoothing techniques described herein.

    Camera Preview Stabilization
    2.
    发明申请

    公开(公告)号:US20250088742A1

    公开(公告)日:2025-03-13

    申请号:US18621956

    申请日:2024-03-29

    Applicant: Apple Inc.

    Abstract: Stabilizing camera preview frames involves obtaining preview frames and associated camera motion signals, smoothing out the trajectory of these frames to correct for camera movement, and then displaying the corrected frames. The process includes comparing current and prior preview frames and their motion signals to calculate smoothed trajectories and apply appropriate rotation and translation corrections. The specifics of trajectory smoothing take into account factors such as motion state, lighting, or capture mode, which can influence the strength of the smoothing parameter. Transform matrices are utilized based on correction rotation to adjust sample points within images, and special considerations like camera sag are factored into the image corrections.

    Method for Camera Alignment Mitigations for Systems with Multiple Cameras

    公开(公告)号:US20250088731A1

    公开(公告)日:2025-03-13

    申请号:US18621999

    申请日:2024-03-29

    Applicant: Apple Inc.

    Abstract: Mitigations for camera alignment during image capture preview involve receiving a zoom level request for a multicamera system comprising a first camera having a first optical zoom metric and a second camera having a second optical zoom metric, and obtaining an image frame from the first camera. A target transformation is determined based on difference between the first optical zoom and a requested zoom level, the second camera's native optical zoom level, the relative orientation of the first and second cameras within the system framework and the relative intrinsics on the two cameras. The transformation is applied before, in conjunction with, or followed by a separate digital zoom transform to the image frame to obtain a transformed and digitally zoomed image frame transformation.

    Distortion compensation in video image stabilization

    公开(公告)号:US12052498B2

    公开(公告)日:2024-07-30

    申请号:US17805119

    申请日:2022-06-02

    Applicant: Apple Inc.

    CPC classification number: H04N23/683 G06T5/80 H04N23/67 H04N23/687

    Abstract: Devices, methods, and non-transitory program storage devices are disclosed to provide techniques for distortion compensation in images that are subject to video image stabilization (VIS), e.g., electronic image stabilization (EIS). In some embodiments, the techniques comprise: obtaining a first image captured by a first image capture device; determining a first set of image parameters configured to apply a first distortion operation (e.g., an un-distortion operation) to the first image; determining a second set of image parameters configured to apply a first stabilization operation to a first version of the first image that has already had the first distortion operation applied; determining a third set of image parameters configured to apply a second distortion operation (e.g., an idealized, field of view-preserving, re-distortion operation) to the first image; and applying the determined first, second, and third sets of image parameters to the first image to generate a distortion-compensated and stabilized output image.

    Distortion Compensation in Video Image Stabilization

    公开(公告)号:US20220394180A1

    公开(公告)日:2022-12-08

    申请号:US17805119

    申请日:2022-06-02

    Applicant: Apple Inc.

    Abstract: Devices, methods, and non-transitory program storage devices are disclosed to provide techniques for distortion compensation in images that are subject to video image stabilization (VIS), e.g., electronic image stabilization (EIS). In some embodiments, the techniques comprise: obtaining a first image captured by a first image capture device; determining a first set of image parameters configured to apply a first distortion operation (e.g., an un-distortion operation) to the first image; determining a second set of image parameters configured to apply a first stabilization operation to a first version of the first image that has already had the first distortion operation applied; determining a third set of image parameters configured to apply a second distortion operation (e.g., an idealized, field of view-preserving, re-distortion operation) to the first image; and applying the determined first, second, and third sets of image parameters to the first image to generate a distortion-compensated and stabilized output image.

Patent Agency Ranking