Abstract: Methods and apparatus for blending unknown pixels in overlapping images. In one embodiment, an action camera captures two hyper-hemispherical fisheye images that are stitched to a 360° panorama. In order to remove exposure differences between the two cameras, the images are pre-processed prior to multiband blending. The pre-processing leverages image information from pixels to make informed guesses about pixels that were not captured. In particular, various pixels with different knowability (e.g., known, unknown, consistent, and conflicting) may be handled differently so as to emphasize/de-emphasize their importance in pre-processing.

Abstract: Methods and apparatus for blending unknown pixels in overlapping images. In one embodiment, an action camera captures two hyper-hemispherical fisheye images that are stitched to a 360° panorama. In order to remove exposure differences between the two cameras, the images are pre-processed prior to multiband blending. The pre-processing leverages image information from pixels to make informed guesses about pixels that were not captured. In particular, various pixels with different knowability (e.g., known, unknown, consistent, and conflicting) may be handled differently so as to emphasize/de-emphasize their importance in pre-processing.

Abstract: Adaptive acquisition control includes obtaining a processed image by an image capture apparatus, which includes, a target exposure component that obtains a target exposure value in accordance with target exposure input data, an aggregate gain component that obtains a target aggregate gain value and a remaining gain value in accordance with aggregate gain input data, an auto-exposure compensation component that obtains an auto-exposure compensation tone curve in accordance with auto-exposure compensation input data, a contrast control component that obtains a contrast control tone curve and a contrast control black point value, and a tone control driver that obtains a tone control tone curve and a tone control black point value, and an image signal processor that processes the current input image in accordance with the tone control tone curve and the tone control black point value to produce the processed image.

Abstract: Lens mode auto-detection includes obtaining predicate lens mode data, obtaining probable lens mode data, obtaining a lens mode score in accordance with the predicate lens mode data and the probable lens mode data, determining whether the lens mode score is greater than a defined lens mode change threshold, and, in response to determining that the lens mode score is greater than the defined lens mode change threshold, outputting lens mode data.

Abstract: A set of image portions is obtained. A pixel series that includes respective representative values of the image portions is obtained. Spectral data are obtained based on the pixel series. Significant local maxima are identified in the spectral data. An artificial light is determined to have been captured in the set of the image portions in response to determining that a number of the significant local maxima is greater than one. An anti-flicker feature of an image capture device is enabled. in response to determining that the artificial light is captured in the set of the image portions.

Abstract: Adaptive acquisition control includes obtaining a processed image by an image capture apparatus, which includes, a target exposure component that obtains a target exposure value in accordance with target exposure input data, an aggregate gain component that obtains a target aggregate gain value and a remaining gain value in accordance with aggregate gain input data, an auto-exposure compensation component that obtains an auto-exposure compensation tone curve in accordance with auto-exposure compensation input data, a contrast control component that obtains a contrast control tone curve and a contrast control black point value, and a tone control driver that obtains a tone control tone curve and a tone control black point value, and an image signal processor that processes the current input image in accordance with the tone control tone curve and the tone control black point value to produce the processed image.

Abstract: Systems and methods are disclosed for denoising chrominance channels of images. For example, methods may include receiving an image from one or more image sensors; determining a set of weights for the image based on a luminance channel of the image, wherein a weight in the set of weights corresponds to a subject pixel and a candidate pixel and is determined based on luminance values of one or more pixels of the image centered at the subject pixel and one or more pixels of the image centered at the candidate pixel; applying the set of weights to chrominance channels of the image to obtain a denoised image, wherein the subject pixel of the denoised image is determined based on the weight multiplied by the candidate pixel of the image; and storing, displaying, or transmitting an output image based on the denoised image.

Abstract: Lens mode auto-detection includes obtaining predicate lens mode data, obtaining probable lens mode data, obtaining a lens mode score in accordance with the predicate lens mode data and the probable lens mode data, determining whether the lens mode score is greater than a defined lens mode change threshold, and, in response to determining that the lens mode score is greater than the defined lens mode change threshold, outputting lens mode data.

Abstract: Optical center calibration may include obtaining one or more parameters for optical center calibration, obtaining an input image captured by an image capture device using a lens, and determining a calibration circle using the parameters and the input image. Determining the calibration circle may include extracting rays using the input image, estimating contours using the input image and the rays, and estimating the calibration circle using the input image and the contours. The calibration may be iteratively improved by repeating calibration based on the input image and a previous iteration of optical center calibration.

Abstract: A video may include multiple video frames. The video frames may be scored based on values of multiple content metrics for individual frames. One or more portions of the video that includes a threshold number of consecutive video frames that meet a score threshold may be identified. For individual ones of the identified portion(s), a video frame may be selected based on a maximum of the score for presentation as an exemplar image.

Abstract: Image analysis and processing may include using an image processor to receive image data corresponding to an input image, determine an initial gain value for the image data based on at least one of a two-dimensional gain map or a parameterized radial gain model, determine whether the initial gain value is below a threshold, determine a maximum RGB triplet value for the image data where the initial gain value is below the threshold, determine a pixel intensity as output of a function for saturation management, determine a final gain value for the image data based on the maximum RGB triplet value and the pixel intensity, apply the final gain value against the image data to produce processed image data, and output the processed image data for further processing using the image processor.

Abstract: Systems and methods are disclosed for denoising chrominance channels of images. For example, methods may include receiving an image from one or more image sensors; determining a set of weights for the image based on a luminance channel of the image, wherein a weight in the set of weights corresponds to a subject pixel and a candidate pixel and is determined based on luminance values of one or more pixels of the image centered at the subject pixel and one or more pixels of the image centered at the candidate pixel; applying the set of weights to chrominance channels of the image to obtain a denoised image, wherein the subject pixel of the denoised image is determined based on the weight multiplied by the candidate pixel of the image; and storing, displaying, or transmitting an output image based on the denoised image.

Abstract: Captured images of a scene may include depictions of objects moving within the scene. The portions of the images depicting the moving objects may be identified by aligning the images and analyzing the changes in pixel values of the aligned images. For the portion of the images depicting the moving objects, the pixels values may be replaced with mean, mode, and/or median values that approximate the value that would have been captured without the moving objects, and one or more image without the depiction of moving objects may be generated.

Abstract: Systems and methods are disclosed for non-local means denoising of images. For example, methods may include receiving an image from an image sensor; determining a set of non-local means weights for the image; applying the set of non-local means weights to the image to obtain a first denoised image; applying the set of non-local means weights to the first denoised image to obtain a second denoised image; and storing, displaying, or transmitting an output image based on the second denoised image.

Abstract: Optical center calibration may include obtaining one or more parameters for optical center calibration, obtaining an input image captured by an image capture device using a lens, and determining a calibration circle using the parameters and the input image. Determining the calibration circle may include extracting rays using the input image, estimating contours using the input image and the rays, and estimating the calibration circle using the input image and the contours. The calibration may be iteratively improved by repeating calibration based on the input image and a previous iteration of optical center calibration.

Abstract: Various embodiments relate generally to data science and data analysis, computer software and systems, and data-driven control systems and algorithms based on graph-based data arrangements, among other things, and, more specifically, to a computing platform configured to receive or analyze datasets in parallel by implementing, for example, parallel computing processor systems to correlate subsets of parallelized data from disparately-formatted data sources to identify entity data and to aggregate graph data portions. In some examples, a method may include classifying data parallelized data to identify a class of observation data, constructing one or more content graphs in a graph data format, correlating parallelized data to other subsets of parallelized data associated with a class of observation data; and aggregating observation data to represent an individual entity.

Abstract: Systems and methods are disclosed for denoising chrominance channels of images. For example, methods may include receiving an image from one or more image sensors; determining a set of weights for the image based on a luminance channel of the image, wherein a weight in the set of weights corresponds to a subject pixel and a candidate pixel and is determined based on luminance values of one or more pixels of the image centered at the subject pixel and one or more pixels of the image centered at the candidate pixel; applying the set of weights to chrominance channels of the image to obtain a denoised image, wherein the subject pixel of the denoised image is determined based on the weight multiplied by the candidate pixel of the image; and storing, displaying, or transmitting an output image based on the denoised image.

Abstract: Methods and apparatus for blending unknown pixels in overlapping images. In one embodiment, an action camera captures two hyper-hemispherical fisheye images that are stitched to a 360° panorama. In order to remove exposure differences between the two cameras, the images are pre-processed prior to multiband blending. The pre-processing leverages image information from pixels to make informed guesses about pixels that were not captured. In particular, various pixels with different knowability (e.g., known, unknown, consistent, and conflicting) may be handled differently so as to emphasize/de-emphasize their importance in pre-processing.

Abstract: Systems and methods are disclosed for denoising chrominance channels of images. For example, methods may include receiving an image from one or more image sensors; determining a set of weights for the image based on a luminance channel of the image, wherein a weight in the set of weights corresponds to a subject pixel and a candidate pixel and is determined based on luminance values of one or more pixels of the image centered at the subject pixel and one or more pixels of the image centered at the candidate pixel; applying the set of weights to chrominance channels of the image to obtain a denoised image, wherein the subject pixel of the denoised image is determined based on the weight multiplied by the candidate pixel of the image; and storing, displaying, or transmitting an output image based on the denoised image.

Abstract: A video may include multiple video frames. The video frames may be scored based on values of multiple content metrics for individual frames. One or more portions of the video that includes a threshold number of consecutive video frames that meet a score threshold may be identified. For individual ones of the identified portion(s), a video frame may be selected based on a maximum of the score for presentation as an exemplar image.