Disclosed is a computer-implemented method of overlaying a representation of a medical instrument over a two-dimensional medical image. It finds at least one feature point along a detection line which is defined relative to the medical instrument in the medical image, calculates a geometrical quantity based on this feature point and adds the geometrical quantity to the two-dimensional medical image.

Images perceived to be substantially full color or multi-colored may be formed using component color images that are distributed in unequal numbers across a plurality of depth planes. The distribution of component color images across the depth planes may vary based on color. In some embodiments, a display system includes a stack of waveguides that each output light of a particular color, with some colors having fewer numbers of associated waveguides than other colors. The stack of waveguides may include by multiple pluralities (e.g., first and second pluralities) of waveguides, each configured to produce an image by outputting light corresponding to a particular color. The total number of waveguides in the second plurality of waveguides is less than the total number of waveguides in the first plurality of waveguides, and may be more than the total number of waveguides in a third plurality of waveguides, in embodiments where three component colors are utilized.

A system and method are provided for capturing an image with correct skin tone exposure. In use, one or more faces having threshold skin tone are detected within a scene. Based on the detected one or more faces, a high dynamic range (HDR) capture mode is enabled. Further, the scene image is captured using the HDR capture mode.

A remote sensing image geometric normalization method and apparatus. The method comprises: constructing a pyramid tile structure for a reference image, and releasing reference tile data, wherein the reference tile data is data in the pyramid tile structure (S11); according to the resolution and geographic coordinates of an image to be subjected to geometric normalization, calculating the level of a tile to be downloaded and the name of the tile to be downloaded, and according to the level of the tile to be downloaded and the name of the tile to be downloaded, downloading corresponding data from the reference tile data to obtain a standard tile set (S12); performing first geometric correction on the image to be subjected to geometric normalization and tiles in the standard tile set to obtain a first image processing result (S13); matching the first image processing result with the tiles in the standard tile set to obtain a plurality of control points, and using the plurality of control points to calculate a result evaluation precision (S14); and according to the result evaluation precision, determining whether to perform second geometric correction on the first image processing result (S15). The method improves the efficiency of processing a remote sensing image.

This disclosure relates to image processing method and apparatus. The method includes: processing, with a first generator in an image processing model, a first sample image in a first sample set to obtain a first predicted image; processing, with the first generator, a second sample image in a second sample set to obtain a second predicted image; and training the image processing model according to a difference between the target avatar in the first sample image and the first predicted avatar and a difference between a first type attribute of the sample avatar in the second sample image and a first type attribute of the second predicted avatar.

Systems and methods are provided for classifying microbial cells according to morphological features of microcolonies. A dark-field objective is employed to acquire a dark-field image of a microcolony during a microcolony growth phase that is characterized by phenotypic expression of microcolony morphological features which evolve with time and are differentiated among classes of microbial cell types. The dark-field image is processed to classify the microcolony according to two or more microbial cell types, such as Gram status and/or speciation. The dark-field objective may have a numerical aperture selected to facilitate the imaging of microcolony morphological features, residing, for example, between 0.15 and 0.35. A set of dark-field images of a microcolony may be collected during the microcolony growth phase and processed to classify the microcolony. Classification may be performed according to a temporal ordering of the dark-field images, for example, using a recurrent neural network.

A noise removing circuit includes an image combiner suitable for generating a high dynamic range (HDR) image by combining images having different exposure times; a detailed image generator suitable for generating a detailed image from the HDR image; an image strength evaluator suitable for evaluating strength of the detailed image; and a noise coring component suitable for performing a noise coring operation for removing noise from a region of the detailed image in which a signal to noise ratio (SNR) has decreased using a low threshold and a saturation threshold when the strength of the detailed image is less than a reference value.

Systems and methods for satellite Synthetic Aperture Radar (SAR) artifact suppression for enhanced three-dimensional feature extraction, change detection, and/or visualizations are described. In some aspects, the described systems and methods include a method for suppressing artifacts from complex SAR data associated with a scene. In some aspects, the described systems and methods include a method for creating a photo-realistic 3D model of a scene based on complex SAR data associated with a scene. In some aspects, the described systems and methods include a method for identifying three-dimensional (3D) features and changes in SAR imagery.

A method and a computing device for object detection and tracking from a video input are described. The method and the computing device may be used to, for example, track objects of interest, such as lane markings, in traffic. A plurality of frames corresponding to a video may be analyzed in a spatiotemporal domain by a neural network. The neural network may be trained using data synthesized in the spatiotemporal domain.

An inter prediction method allows a variable coefficient deep learning model to adaptively learn characteristics of a video; transmits a variable coefficient deep learning model parameter generated from the learning from an image encoding device to an image decoding device; and refers to a virtual reference frame generated by the variable coefficient deep learning model.