摘要:
A covariance matrix associated for a set of spatial and/or temporal sub-portions of a media object is populated based on a lookup table of relative displacements and sums of covariance values generated from pairs of data elements in the media object. The lookup table is generated in linear time, providing an efficient method of generating the covariance matrix.
摘要:
In some instances, an image may have dimensions that do not correspond to a slot to display the image. For example, an image content item may have dimensions that do not correspond to a content item slot. The image may be resized using seam carving to add or remove pixels of the image. A saliency map for the image may be used having saliency scores for each pixel of the image. Evaluation metrics may be used before, during, and after, seam carving to determine whether salient content is affected by the seam carving. In some instances, a seam cost threshold value may be used for adaptive step size during the seam carving. The resized image may then be outputted, such as for an image content item to be served with a resource.
摘要:
An image processing server performs haze-removal from images. Global atmospheric light is estimated and an initial transmission value is estimated. In one embodiment, a solver is applied to an objective function to recover a scene radiance value based on the estimated atmospheric light and estimated transmission value. The scene radiance value is used to construct an image without haze. In a simplified method that avoids using a solver, bilateral filtering is performed on the transmission image in order to construct an image without haze.
摘要:
Methods and systems for rolling shutter removal are described. A computing device may be configured to determine, in a frame of a video, distinguishable features. The frame may include sets of pixels captured asynchronously. The computing device may be configured to determine for a pixel representing a feature in the frame, a corresponding pixel representing the feature in a consecutive frame; and determine, for a set of pixels including the pixel in the frame, a projective transform that may represent motion of the camera. The computing device may be configured to determine, for the set of pixels in the frame, a mixture transform based on a combination of the projective transform and respective projective transforms determined for other sets of pixels. Accordingly, the computing device may be configured to estimate a motion path of the camera to account for distortion associated with the asynchronous capturing of the sets of pixels.
摘要:
Methods and systems for processing a video for stabilization and retargeting are described. A recorded video may be stabilized by removing shake introduced in the video, and a video may be retargeted by modifying the video to fit to a different aspect ratio. Constraints can be imposed that require a modified video to contain pixels from the original video and/or to preserve salient regions. In one example, a video may be processed to estimate an original path of a camera that recorded the video, to estimate a new camera path, and to recast the video from the original path to the new camera path. To estimate a new camera path, a virtual crop window can be designated. A difference transformation between the original and new camera path can be applied to the video using the crop window to recast the recorded video from the smooth camera path.
摘要:
Methods and systems for processing an image to create an object model are disclosed. In accordance with one embodiment, each segment of the image is assigned to a respective bin of a bounding box. For each bin of the bounding box, the value of a feature for the bin is computed based on the values of that feature for each of the segments assigned to the bin. An object model is then created based on the values of the feature for the bin.
摘要:
A video is resized while preserving salient regions within the video. The dimensions of the video are reduced by selectively removing unimportant, or less salient, pixels from the video in a manner that diminishes any resulting spatial and temporal artifacts. More specifically, “seams” of pixels are selectively removed from frames of the video, where a “seam” is a horizontal or vertical chain of pixels through a frame. A seam can be selected from among all possible pixels in a row or column, thereby allowing discontinuous seams. Seams are selected using a technique that encourages the seam to pass through less-salient regions of the frame, and that reduces spatial artifacts when the video is played.