7 results about "Synthetic projection" patented technology

A deep learning-enabled system for the display or projection of high-resolution images is disclosed that is based on a jointly-trained pair of an electronic encoder network and an all-optical decoder network to synthesize / project super-resolved images using low-resolution wavefront modulators. The electronic encoder network rapidly pre-processes the high-resolution images of interest so that their spatial information is encoded into low-resolution (LR) modulation patterns, projected via a low SBP wavefront modulator. The all-optical decoder network processes this LR encoded information using thin transmissive layers that are structured using deep learning to all-optically synthesize and project super-resolved images at its output FOV. Results indicate that this diffractive image display system can achieve a super-resolution factor of ˜4, demonstrating a ˜16-fold increase in SBP. The system can be scaled to operate at visible wavelengths and be used for large FOV and high-resolution displays that are compact, low-power, and computationally efficient.

The present invention generally relates to a technology for recognizing objects in a Around View Monitoring (AVM) video for a vehicle. In particular, the present invention relates to a DNN-based object recognition method for multi-channel fisheye images of a vehicle AVM apparatus, by projecting multi-channel unit images of fisheye-lens cameras of the vehicle AVM apparatus onto a cylindrical projection plane so as to obtain multi-channel unit projection images, combining the multi-channel unit projection images on a single image so as to obtain a composite projection image, and inputting the composite projection image to a Deep Neural Network (DNN) model so as to obtain object recognition output. The invention has an advantage that AVM apparatus can effectively assist in safe driving by monitoring objects around the vehicle. Further, the present invention has an advantage that object recognition may be effectively achieved by the insufficient computing resources of vehicles.

A method for generating a plurality of contrast-enhanced mammograms of an examination region is described. In the method a plurality of at least two different angular positions is defined for a contrast-enhanced tomosynthesis projection recording of an examination region. At least two low-energy tomosynthesis projection recordings of the examination region are subsequently generated, starting from the at least two different angular positions, in the presence of a contrast agent. Furthermore, at least two high-energy tomosynthesis projection recordings of the examination region are generated, starting from the at least two different angular positions, in the presence of a contrast agent. Finally, at least two contrast-enhanced mammograms are generated by subtracting a respective low-energy tomosynthesis projection recording from a respective high-energy tomosynthesis projection recording for a respective angular position. In addition, a mammography system is described.

The present invention provides a multi-window different-display screen projection method based on a container execution mechanism, which relates to the field of smart terminal display technology. The method comprises constructing multiple target containers based on active components and arranging them according to a preset layout template to generate a virtual main window, obtaining a user touch screen event through the virtual main window and converting its original coordinates into target coordinates, and then sending the event to a target application window in the target container, driving the target application window to execute logical operations and generate real-time picture frames to store in a frame buffer of the target container, reading all real-time picture frames through the virtual main window, splicing them according to a preset layout template to obtain a synthetic projection screen, encoding the synthetic projection screen and transmitting it to an external screen projection end through a projection protocol, thereby improving development efficiency, realizing multi-window different-display screen projection, enriching projection diversity, and flexibly arranging a variety of containers in the main virtual window to adapt to different large-screen scenarios and meet diverse projection needs.

This invention relates to the field of image processing technology and discloses a method and apparatus for processing projection data in cone-beam CT half-fan scanning. The method includes: acquiring raw projection data obtained from cone-beam CT half-fan scanning; inputting a first diagonal projection data pair from the raw projection data into a pre-trained artificial intelligence model to obtain a pair of synthetic projection data, wherein the size of the pair of synthetic projection data is larger than the raw projection data and equivalent to the projection data obtained from full-fan scanning. This invention, through AI, learns complete projection data from two basic diagonal (approximately 180° interval) projection data, obtaining projection data equivalent to that acquired by a large-size detector, thereby increasing the scanning field of view of cone-beam CT.

The invention discloses a cone beam CT image reconstruction method and device, equipment and a storage medium. The method comprises the following steps: acquiring limited angle projection data and geometric parameters; inputting the three-dimensional Gaussian kernel parameter into a pre-trained generative model to obtain an initial Gaussian kernel parameter of the three-dimensional Gaussian scene; taking the initial scene as a starting point, carrying out iterative optimization under the constraint of a micro-cone beam CT forward projection model, obtaining a synthetic projection through micro-rendering in each iteration, calculating an error between the synthetic projection and a real projection, updating Gaussian kernel parameters according to the error, and adaptively adjusting the density distribution of Gaussian kernels; and finally, converting the optimized scene into three-dimensional volume data and outputting the three-dimensional volume data. According to the method, strong structure prior initialization is provided through the generative model, and high-quality and high-efficiency reconstruction under the condition of serious data missing is realized in combination with iterative optimization constrained by the physical model and adaptive structure adjustment, so that artifacts are remarkably inhibited, and a clear anatomical structure is recovered.

The invention relates to a virtual shooting content and space adaptation method, which comprises the following steps of: establishing a virtual canvas, and drawing a fusion device and a display area of a display channel thereof on the virtual canvas based on space layout information to form a space mapping model; arranging the virtual shooting content to a target position of the virtual canvas, and generating a content mapping matrix; based on the space mapping model and the content mapping matrix, calculating the serial number, the channel resolution and the pixel proportion of a display channel corresponding to the virtual shooting content at the target position of the virtual canvas; performing image segmentation on the virtual shooting content based on the channel resolution and the pixel ratio; synthesizing the virtual shooting content corresponding to each display channel belonging to the same fusion device to obtain the synthesized projection content of the corresponding fusion device; the fusion device receives the corresponding synthesized projection content and distributes the corresponding display channel for projection output; by the adoption of the technical scheme, the efficiency of editing operation is improved, and multi-channel intelligent adaptation of the virtual content is achieved.