Abstract: Various methods for utilizing a saliency heatmaps are described. The methods include obtaining image data corresponding to an image of a scene, obtaining a saliency heatmap for the image of the scene based on a saliency network, wherein the saliency heatmap indicates a likelihood of saliency for a corresponding portion of the scene, and manipulating the image data based on the saliency heatmap. In embodiments, the saliency heatmap may be produced using a trained machine learning model. The saliency heatmap may be used for various image processing tasks, such as determining which portion(s) of a scene to base an image capture device's autofocus, auto exposure, and/or white balance operations upon. According to some embodiments, one or more bounding boxes may be generated based on the saliency heatmap, e.g., using an optimization operation, which bounding box(es) may be used to assist or enhance the performance of various image processing tasks.

Abstract: Various methods for utilizing a saliency heatmaps are described. The methods include obtaining image data corresponding to an image of a scene, obtaining a saliency heatmap for the image of the scene based on a saliency network, wherein the saliency heatmap indicates a likelihood of saliency for a corresponding portion of the scene, and manipulating the image data based on the saliency heatmap. In embodiments, the saliency heatmap may be produced using a trained machine learning model. The saliency heatmap may be used for various image processing tasks, such as determining which portion(s) of a scene to base an image capture device's autofocus, auto exposure, and/or white balance operations upon. According to some embodiments, one or more bounding boxes may be generated based on the saliency heatmap, e.g., using an optimization operation, which bounding box(es) may be used to assist or enhance the performance of various image processing tasks.

Abstract: Various methods for utilizing a saliency heatmaps are described. The methods include obtaining image data corresponding to an image of a scene, obtaining a saliency heatmap for the image of the scene based on a saliency network, wherein the saliency heatmap indicates a likelihood of saliency for a corresponding portion of the scene, and manipulating the image data based on the saliency heatmap. In embodiments, the saliency heatmap may be produced using a trained machine learning model. The saliency heatmap may be used for various image processing tasks, such as determining which portion(s) of a scene to base an image capture device's autofocus, auto exposure, and/or white balance operations upon. According to some embodiments, one or more bounding boxes may be generated based on the saliency heatmap, e.g., using an optimization operation, which bounding box(es) may be used to assist or enhance the performance of various image processing tasks.

Abstract: An electronic image capture device captures a first image of a scene at a first time. A first local tone mapping operator for a first portion of the first image is determined. The electronic image capture device further captures a second image of the scene at a second time. A motion of the electronic device between the first time and the second time is determined. A second local tone mapping operator for a second portion of the second image is determined. The second portion is determined to correspond to the first portion based, at least in part, on the determined motion of the electronic device. The second local tone mapping operator is determined based, at least in part, on the first local tone mapping operator. At least the second local tone mapping operator is applied to the second portion of the second image.

Abstract: Determining a focus setting includes determining a plurality of regions of interest in a view of a scene, and, for each of the plurality of regions of interest, obtaining a set of image data for each of multiple focal positions, and then applying focus filters to the set of image data for each of the plurality of focal positions for each of the regions of interest to obtain a set of focus scores, i.e., a focus score for each focus filter applied to the set of image data for each of the focal positions. Further, determining a confidence value associated with each of the sets of focus scores, selecting a subset of the sets of focus scores based on the confidence values associated with each of the sets of focus scores, and determining a focus setting for the scene based on the selected subset of the focus scores.

Abstract: Various methods for utilizing a saliency heatmaps are described. The methods include obtaining image data corresponding to an image of a scene, obtaining a saliency heatmap for the image of the scene based on a saliency network, wherein the saliency heatmap indicates a likelihood of saliency for a corresponding portion of the scene, and manipulating the image data based on the saliency heatmap. In embodiments, the saliency heatmap may be produced using a trained machine learning model. The saliency heatmap may be used for various image processing tasks, such as determining which portion(s) of a scene to base an image capture device's autofocus, auto exposure, and/or white balance operations upon. According to some embodiments, one or more bounding boxes may be generated based on the saliency heatmap, e.g., using an optimization operation, which bounding box(es) may be used to assist or enhance the performance of various image processing tasks.

Abstract: Techniques for auto white balancing of captured images based on detection of flicker in ambient light is described. When flicker is detected in ambient light during an image capture event, and the flicker is unchanging during the image capture event, a white point of image data may be estimated according to a first technique. When flicker is detected in ambient light during an image capture event, and the flicker is changing during the image capture event, a white point of image data may be estimated according to a second technique. When flicker is not detected, a white point of image data may be estimated according to a third technique. Image data may be color corrected based on the estimated white point.

Abstract: Systems and methods are disclosed for correcting red-eye artifacts in a target image of a subject. Images, captured by a camera, including a raw image, are used to generate the target image. An eye region of the target image is modulated to correct for the red-eye artifacts, wherein correction is carried out based on information extracted from at least one of the raw image and the target image. Modulation comprises detecting landmarks associated with the eye region; estimating spectral response of the red eye artifacts; segmenting an image region of the eye based on the estimated spectral response of the red eye artifacts and the detected landmarks, forming a repair mask; and modifying an image region associated with the repair mask.

Abstract: Techniques for adjusting formats of images and video are presented, for example where an SDR source is presented on an HDR display, or vice versa. Techniques include deriving a conversion profile for image data where the conversion profile is responsive to parameters describing characteristics of a domain of source image data and characteristics of a domain of a target rendering environment. Some techniques include creating a tone curve from weighted basis functions.

Abstract: This disclosure relates to a wide gamut encoder capable of receiving a wide gamut color image in accordance with a wide gamut standard. The encoder can encode one or more wide gamut color image pixel values into portions of narrow gamut encoding elements for transmission to a video encoder. The encoder can implement an advanced extended YCC format that is backward compatible with a P3 color gamut.

Abstract: This disclosure pertains to systems, methods, and computer readable media for performing lens shading correction (LSC) operations that modulate gains based on scene lux level and lens focus distance. These gains compensate for both color lens shading (i.e., the deviation between R, G, and B channels) and vignetting (i.e., the drop off in pixel intensity around the edges of an image). As scene illuminance increases, the sensor captures more signal from the actual scene, and the lens shading effects begin to appear. To deal with the situation, the lens shading gains are configured to adaptively ‘scale down’ when scene lux approaches zero and ‘scale up’ when scene lux changes from near zero to become larger. The lens shading gain may also be modulated based on the focus distance. For optical systems without zoom, the inventors have discovered that the amount of lens shading fall off changes as focus distance changes.

Abstract: Displaying wide-gamut images as intended on color-managed wide-gamut display systems while rendering a visually consistent image when rendered on targeted narrow-gamut display systems (regardless of whether the narrow-gamut displays are color-managed). Images represented in accordance with this disclosure are referred to as a dual-target images (DTI): one target being the image's original wide-gamut color space, the other target being a specified narrow-gamut color space. The novel representational scheme retains narrow-gamut rendering for those colors in a wide-gamut image that are within the targeted narrow-gamut color space, transitioning to wide-gamut rendering for those colors in the wide-gamut image that are outside the targeted narrow-gamut color space. This approach minimizes pixel clipping when rendering a wide-gamut image for a narrow-gamut display, while allowing the original wide-gamut pixel values to be recovered when rendering for a wide-gamut display.

Abstract: An image system with a dual layer photodiode structure is provided for processing color images. In particular, the image system can include an image sensor that can include photodiodes with a dual layer photodiode structure. In some embodiments, the dual layer photodiode can include a first layer of photodiodes (e.g., a bottom layer), an insulation layer disposed on the first layer of photodiodes, and a second layer of photodiodes (e.g., a top layer) disposed on the insulation layer. The first layer of photodiodes can include one or more suitable pixels (e.g., green, blue, clear, luminance, and/or infrared pixels). Likewise, the second layer of photodiodes can include one or more suitable pixels (e.g., green, red, clear, luminance, and/or infrared pixels). An image sensor incorporating dual layer photodiodes can gain light sensitivity with additional clear pixels and maintain luminance information with green pixels.

Abstract: A gamut size metric is used in all phases of color image processing (e.g., capture, transmission, and display). In general, the gamut size metric is a single-valued metric that changes as image content changes. More particularly, a gamut boundary histogram is determined and used to estimate a gamut size metric. A gamut size metric identifies a minimum size gamut needed to encompass each pixel in an image, where the gamut size is limited at one end by a first device independent gamut (S1), and at another end by a second device independent color space (S2), where S1 is wholly enclosed within S2. The gamut size metric may be based on strict pixel color value differences. In other embodiments the gamut size metric may take into effect perceptual color differences and significance.

Abstract: Techniques for auto white balancing of captured images based on detection of flicker in ambient light is described. When flicker is detected in ambient light during an image capture event, and the flicker is unchanging during the image capture event, a white point of image data may be estimated according to a first technique. When flicker is detected in ambient light during an image capture event, and the flicker is changing during the image capture event, a white point of image data may be estimated according to a second technique. When flicker is not detected, a white point of image data may be estimated according to a third technique. Image data may be color corrected based on the estimated white point.

Abstract: Methods, devices and computer readable instructions to generate region-of-interest (ROI) tone curves are disclosed. One method includes obtaining a statistic for an entire image such as, for example, a luminance statistic. The same statistic may then be found for a specified ROI of the image. A weighted combination of the statistic of the entire image and the statistic of the ROI yields a combined statistic which may then be converted to a ROI-biased tone curve. The weight used to combine the two statistics may be selected to emphasize or de-emphasize the role of the ROI's statistic in the final tone curve.

Abstract: This disclosure pertains to systems, methods, and computer readable media for performing lens shading correction (LSC) operations that modulate gains based on scene lux level and lens focus distance. These gains compensate for both color lens shading (i.e., the deviation between R, G, and B channels) and vignetting (i.e., the drop off in pixel intensity around the edges of an image). As scene illuminance increases, the sensor captures more signal from the actual scene, and the lens shading effects begin to appear. To deal with the situation, the lens shading gains are configured to adaptively ‘scale down’ when scene lux approaches zero and ‘scale up’ when scene lux changes from near zero to become larger. The lens shading gain may also be modulated based on the focus distance. For optical systems without zoom, the inventors have discovered that the amount of lens shading fall off changes as focus distance changes.

Abstract: This disclosure relates to a wide gamut encoder capable of receiving a wide gamut color image in accordance with a wide gamut standard. The encoder can encode one or more wide gamut color image pixel values into portions of narrow gamut encoding elements for transmission to a video encoder. The encoder can implement an advanced extended YCC format that is backward compatible with a P3 color gamut.

Abstract: This disclosure pertains to systems, methods, and computer readable media for performing lens shading correction (LSC) operations that modulate gains based on scene lux level and lens focus distance. These gains compensate for both color lens shading (i.e., the deviation between R, G, and B channels) and vignetting (i.e., the drop off in pixel intensity around the edges of an image). As scene illuminance increases, the sensor captures more signal from the actual scene, and the lens shading effects begin to appear. To deal with the situation, the lens shading gains are configured to adaptively ‘scale down’ when scene lux approaches zero and ‘scale up’ when scene lux changes from near zero to become larger. The lens shading gain may also be modulated based on the focus distance. For optical systems without zoom, the inventors have discovered that the amount of lens shading fall off changes as focus distance changes.

Abstract: In general, techniques are disclosed for displaying wide-gamut images as intended on color-managed wide-gamut display systems while rendering a visually consistent image when rendered on targeted narrow-gamut display systems (regardless of whether the narrow-gamut displays are color-managed). For this reason, an image represented in accordance with this disclosure is referred to as a dual-target image (DTI): one target being the image's original wide-gamut color space, the other target being a specified narrow-gamut color space. The novel representational scheme described herein retains narrow-gamut rendering for those colors in a wide-gamut image that are within the targeted narrow-gamut color space, transitioning to wide-gamut rendering for those colors in the wide-gamut image that are outside the targeted narrow-gamut color space.