The present disclosure provides a method of processing an image, a device, and a medium. The method of processing the image includes: performing a noise reduction on an original image to obtain a smooth image; performing a feature extraction on the original image to obtain feature data for at least one direction; and determining an image quality of the original image according to the original image, the smooth image, and the feature data for the at least one direction.

Accurate tissue segmentation is performed without a priori knowledge of tissue type or other extrinsic information not found within the subject image, and may be combined with classification analysis so that diseased tissue is not only delineated within an image but also characterized in terms of disease type. In various embodiments, a source image is decomposed into smaller overlapping subimages such as square or rectangular tiles. A predictor such as a convolutional neural network produces tile-level classifications that are aggregated to produce a tissue segmentation and, in some embodiments, to classify the source image or a subregion thereof.

A method is described. The method includes receiving oversampled infrared data provided from an infrared pixel array. The method also includes performing at least one of selective median filtering, spatial-temporal filtering, or resolution enhancement for the oversampled infrared data.

A method for generating a color-faithful extended-depth-of-field (EDF) image from a color volume of 2D images acquired at different focal depths using a microscope. The method involves: generating a grayscale volume; applying invertible color-to-grayscale transformation to the volume; applying wavelet transform to the grayscale volume to obtain a 3D wavelet-coefficient-matrix (WCM); selecting wavelet coefficients using a coefficient selection rule; generating a 2D-WCM and a 2D coefficient-map (CM); applying inverse transformation of the wavelet transform to the 2D-WCM to obtain a 2D grayscale EDF image; generating a 2D color-composite(CC) image; applying inverse transformation of the color-to-grayscale transformation to the 2D grayscale EDF image to obtain a 2D color EDF image; converting the 2D-CC image and the 2D color EDF image into a color space including chromaticity and intensity component(s); and concatenating, chromaticity component(s) of the 2D-CC image and intensity component(s) of the 2D color EDF image, to obtain a color-faithful EDF image.

In one embodiment, a method includes accessing a plurality of image frames captured by one or more cameras, classifying one or more first objects detected in one or more first image frames of the plurality of image frames as undesirable, applying a pixel filtering to the one or more first image frames to replace one or more first pixel sets associated with the one or more first objects with pixels from one or more second image frames of the plurality of image frames to generate a final image frame, providing the final image frame for display.

Systems and methods for hybrid depth regularization in accordance with various embodiments of the invention are disclosed. In one embodiment of the invention, a depth sensing system comprises a plurality of cameras; a processor; and a memory containing an image processing application. The image processing application may direct the processor to obtain image data for a plurality of images from multiple viewpoints, the image data comprising a reference image and at least one alternate view image; generate a raw depth map using a first depth estimation process, and a confidence map; and generate a regularized depth map. The regularized depth map may be generated by computing a secondary depth map using a second different depth estimation process; and computing a composite depth map by selecting depth estimates from the raw depth map and the secondary depth map based on the confidence map.

An image detection obtains original image. The original image is corrected to obtain a corrected image. Median filtering is performed on the corrected image to obtain a filtered image. A contrast of the filtered image is adjusted to obtain an adjusted image. Bilateral filtering is performed on the adjusted image to obtain an enhanced image. Defects in the enhanced image is detected. The method can detect defects in images accurately and efficiently.

The present invention discloses a method and device for automatically drawing structural cracks and precisely measuring widths thereof. The method comprises a method for automatically drawing cracks and a method for calculating widths of these cracks based on a single-pixel skeleton and Zernike orthogonal moments, wherein the method for automatically drawing cracks is used to rapidly and precisely draw cracks in the surface of a structure, and the method for calculating widths of these cracks based on a single-pixel skeleton and Zernike orthogonal moments is used to calculate widths of macro-cracks and micro-cracks in an image in a real-time manner.

A pupil tracking method includes: retrieving an external eye image of a subject, wherein the external eye image includes a pupil of the subject; performing an image preprocessing on the external eye image, wherein the image preprocessing includes performing a binary conversion on the external eye image to obtain a binary image; finding out a contour boundary of each feature in the binary image, and finding out a pupil feature based on a variance of a distance from the contour boundary of each feature to a corresponding reference point; fitting the contour boundary of the pupil feature by a boundary fitting method to find a center coordinate of the pupil feature. The abovementioned pupil tracking method can track the pupil of the subject's eyeball without using a stereo camera. An ophthalmology inspection device using the abovementioned pupil tracking method is also disclosed.

Apparatuses, systems, and techniques to calculate motion of one or more pixel in a region in an image. In at least one embodiment, motion is calculated based on motion of one or more pixels in a different region of said image that overlaps said region in which one or more algorithms are expressed in CUDA code, for example, for efficient execution on a GPU.