摘要:
A system for encoding a video stream into a processed video signal that includes at least one image, includes a pattern detection module for detecting a pattern of interest in the at least one image and identifying a region that contains the pattern of interest when the pattern of interest is detected, based on an analysis of the image in a plurality of domains. An encoder section, generates the processed video signal and wherein, when the pattern of interest is detected, a higher quality, such as a higher bit allocation or higher computational processing, is assigned to the region than to portions of the at least one image outside the region.
摘要:
A transcoder includes a direct transcoder that generates a first portion of a transcoded video stream by reusing a plurality of encoding parameters of a compressed video stream. A cascaded transcoder generates a second portion of the transcoded video stream by decoding the compressed video stream into video data in an uncompressed video format and by re-encoding the video data. A transcoding decision generator generates a transcoding indicator, based on the compressed video stream. A switching module selects the direct transcoder for the first portion of the transcoded video stream and the cascaded transcoder for the second portion of the transcoded video stream, based on the transcoding indicator.
摘要:
A two-dimensional filter operates in accordance with a three-dimensional television signal having a basic layer, a reference processing unit layer and an enhancement layer. The two-dimensional filter generates a processed base layer reference picture from a base layer reference picture via a one-dimensional horizontal filter that horizontally filters the base layer reference picture to generate a first filtered base layer reference picture and a one-dimensional vertical filter, coupled to, but separable from, the one-dimensional horizontal filter, that vertically filters the first filtered base layer reference picture to generate a pre-processed base layer reference picture.
摘要:
A two-dimensional filter operates in accordance with a three-dimensional television signal having a basic layer, a reference processing unit layer and an enhancement layer. The two-dimensional filter generates a processed base layer reference picture from a base layer reference picture via a one-dimensional horizontal filter that horizontally filters the base layer reference picture to generate a first filtered base layer reference picture and a one-dimensional vertical filter, coupled to, but separable from, the one-dimensional horizontal filter, that vertically filters the first filtered base layer reference picture to generate a pre-processed base layer reference picture.
摘要:
A video filter includes a filter parameter generator receives a non-quantization coding parameter corresponding to video data and generates a plurality of deblocking filter parameters in response thereto. An adaptive deblocking filter filters the video data to generate processed video data, based on the plurality of deblocking filter parameters.
摘要:
A system for encoding a video stream into a processed video signal that includes at least one image, includes a region identification signal generator for detecting a region of interest in the at least one image and generating a region identification signal when the pattern of interest is detected. An encoder section generates the processed video signal based on the operation of a plurality of encoding tools, each having at least one encoder quality parameter. The encoder section adjusts the at least one encoding quality parameter of at least one of the plurality of encoding tools in response to the region identification signal.
摘要:
An imaging and display system for guiding medical interventions includes a wearable display, such as a goggle display, for viewing by a user. The display presents a composite, or combined image that includes pre-operative surgical navigation images, intraoperative images, and in-vivo microscopy images or sensing data. The pre-operative images are acquired from scanners, such as MRI and CT scanners, while the intra-operative images are acquired in real-time from a camera system carried by the goggle display for imaging the patient being treated so as to acquire intraoperative images, such as fluorescence images. A probe, such as a microscopy probe or a sensing probe, is used to acquire in-vivo imaging/sensing data from the patient. Additionally, the intra-operative and in-vivo images are acquired using tracking and registration techniques to align them with the pre-operative image and the patient to form a composite image for display by the goggle display.
摘要:
A system for using radio frequency (RF) communication signals to extract situational awareness information thorough deep learning. Pre-processing is performed to maximally preserve discriminating features in spatial, temporal and frequency domains. A specially designed neural network architecture is used for handling complex RF signals and extracting spatial, temporal and frequency domain information. Data collection and training is used so that the learning system desensitizes from features orthogonal to the underlying classification problem.