A computer vision system is provided that includes an image generation device configured to capture consecutive two dimensional (2D) images of a scene, a first memory configured to store the consecutive 2D images, a second memory configured to store a growing window of consecutive rows of a reference image and a growing window of consecutive rows of a current image, wherein the reference image and the current image are a pair of consecutive 2D images stored in the first memory, a third memory configured to store a sliding window of pixels fetched from the growing window of the reference image, wherein the pixels in the sliding window are stored in tiles, and a dense optical flow engine (DOFE) configured to determine a dense optical flow map for the pair of consecutive 2D images, wherein the DOFE uses the sliding window as a search window for pixel correspondence searches.

Systems, methods, and apparatuses for generating a subsurface image using diffraction energy information are disclosed. The systems, methods, and apparatuses may include converting a shot gather into one or more plane-wave gather using a Radon transform. The plane-wave gathers may be extrapolated into source-side wavefields and receiver-side wavefields and further generate a pseudo dip-angle gather. The diffraction energy information may be extracted from the pseudo dip-angle gather, and an image containing subsurface features may be generated from the extracted diffraction energy information. The receiver-side wavefields may be decomposed using a recursive Radon transform.

An inspection method according to one aspect of the present disclosure, includes: acquiring a plurality of first acquisition images of a first die having a predetermined pattern; acquiring a plurality of second acquisition images of a second die having the predetermined pattern; acquiring a plurality of third acquisition images of a third die having the predetermined pattern; producing a first average image with the first acquisition images; producing a second average image with the second acquisition images: producing a third average image with the third acquisition images; producing a first comparative image with the first average image and the second average image; producing a second comparative image with the second average image and the third average image; producing a third comparative image with the first average image and the third average image; producing a reference image with the first acquisition images, the second acquisition images, and the third acquisition images; performing first comparison between the first comparative image and the reference image; performing second comparison between the second comparative image and the reference image; performing third comparison between the third comparative image and the reference image; determining, when a first defect is detected in the first comparison and the third comparison, that the first die has the first defect; determining, when a second defect is detected in the first comparison and the second comparison, that the second die has the second defect; and determining, when a third defect is detected in the second comparison and the third comparison, that the third die has the third defect.

A quantitative statistical characterization method of micron-level second phase in aluminum alloy based on deep learning is disclosed. The method includes obtaining a feature database of the standard sample, training the feature database by the image segmentation network U-Net based on deep learning to obtain a U-Net segmentation model, selecting the corresponding parameters of the optimal precision and establishing a U-Net target model; clipping the aluminum alloy image to be detected and inputting the clipped images into the U-net target model, obtaining the size, area and position information of the second phase through the connected region algorithm, carrying out statistical distribution of the data set combined with the mathematical statistical method, and restoring the position information to the surface of the aluminum alloy to be tested to obtain the full-field quantitative statistical distribution and visualization results.

A method for reducing image disturbances caused by reconstructed defective pixel clusters located in signal-gradient affected diagnostic image regions. An individually adapted central symmetrical pair reconstruction (CSP) kernel is composed for a defective image pixel based on a kernel-pair candidate order encoded in a model thereby using the pixel's validity state. The image impacted by defective pixels is corrected in real-time by statistical filtering or spatial convolution of the kernel-associated image data accessible via a predetermined CSP kernels image-offsets structure.

The present disclosure relates to a simple, fast, and low cost method for 3D microvascular imaging, termed “scatter labeled imaging of microvasculature in excised tissue” (SLIME). The method can include perfusing a contrast agent through vasculature of a tissue sample. The contrast agent can include colloids and a dispersant. After the contrast agent is perfused through the vasculature, the vasculature of the tissue sample can be treated with a molecule that cross links with at least a portion of the dispersant to form a sticky, non-Newtonian polymer that prevents leakage of the contrast agent out of the vasculature of the tissue sample. The tissue sample can then be immersed in a solution comprising a clearing agent and subsequently imaged.

The present application discloses a method of adjusting a parameter, the parameter being used to derive a physiological characteristic of an individual from an image of the user, the method comprising the steps of: obtaining the parameter for the individual; obtaining a corresponding parameter for a plurality of other individuals within a cohort of the individual; comparing the parameter for the individual with a statistically significant parameter for the plurality of other individuals; and adjusting the parameter for the individual in accordance with the difference between the parameter for the individual and the statistically significant parameter for the plurality of other individuals.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for correcting a corrupted data sample using a trained deep neural network, the method including obtaining a feature representation of a corrupted data sample; and modifying the feature representation of the corrupted data sample to generate a feature representation of a corrected data sample by iteratively processing a current version of the feature representation of the corrupted data sample using the trained deep neural network to generate a current corruption score for the current version of the feature representation of the corrupted data sample and generating a less-corrupted version of the feature representation by performing an iteration of gradient descent against the current version of the feature representation of the corrupted data sample to reduce the current corruption score.

In part, the disclosure relates to an automated method of branch detection with regard to a blood vessel imaged using an intravascular modality such as OCT, IVUS, or other imaging modalities. In one embodiment, a representation of A-lines and frames generated using an intravascular imaging system is used to identify candidate branches of a blood vessel. One or more operators such as filters can be applied to remove false positives associated with other detections.

An image-filtering device for filtering an image that includes a pixel difference computing module, an adaptive brightness adjusting module, a weighting computing module and a filter computing module is provided. The pixel difference computing module uses any one of the pixels as a central pixel within a pixel window to compute pixel absolute differences between the central pixel and every pixels within the pixel window. The adaptive brightness adjusting module multiplies each of the pixel absolute differences with an adjusting parameter to generate adjusted pixel absolute differences. The weighting computing module generates weighting values according to the adjusted pixel absolute differences. The filter computing module performs convolution according to the pixel value of each of the pixels within the pixel window and the corresponding weighting values to generate a filtering result of the central pixel.