摘要:
A method detects short term, unusual events in a video. First, features are extracted features from the audio and the video portions of the video. Segments of the video are labeled according to the features. A global sliding window is applied to the labeled segments to determine global characteristics over time, while a local sliding window is applied only to the labeled segments of the global sliding window to determine local characteristic over time. The local window is substantially shorter in time than the global window. A distance between the global and local characteristic is measured to determine occurrences of the unusual short time events.
摘要:
A method identifies highlight segments in a video including a sequence of frames. Audio objects are detected to identify frames associated with audio events in the video, and visual objects are detected to identify frames associated with visual events. Selected visual objects are matched with an associated audio object to form an audio-visual object only if the selected visual object matches the associated audio object, the audio-visual object identifying a candidate highlight segment. The candidate highlight segments are further refined, using low level features, to eliminate false highlight segments.
摘要:
A method presents a video according to compositional structures associated with the video. Each compositional structure has a label, and multiple segments that can be organized temporally or hierarchically. A particular compositional structure is selected with a remote controller, and the video is presented by a playback controller on a display device according to the compositional structure.
摘要:
A method uses probabilistic fusion to detect highlights in videos using both audio and visual information. Specifically, the method uses coupled hidden Markov models (CHMMs). Audio labels are generated using audio classification via Gaussian mixture models (GMMs), and visual labels are generated by quantizing average motion vector magnitudes. Highlights are modeled using discrete-observation CHMMs trained with labeled videos. The CHMMs have better performance than conventional hidden Markov models (HMMs) trained only on audio signals, or only on video frames.
摘要:
An auto-commissioning system provides automatic parameter selection for an intelligent video system based on target video provided by the intelligent video system. The auto-commissioning system extracts visual feature descriptors from the target video and provides the one or more visual feature descriptors associated with the received target video to an parameter database that is comprised of a plurality of entries, each entry including a set of one or more stored visual feature descriptors and associated parameters tailored for the set of stored visual feature descriptors. A search of the parameter database locates one or more best matches between the extracted visual feature descriptors and the stored visual feature descriptors. The parameters associated with the best matches are returned as part of the search and used to commission the intelligent video system.
摘要:
A system and method for performing automated defect detection by utilizing data from prior inspections is disclosed. The system and method may include providing a image capture device for capturing and transmitting at least one current image of an object and providing a database for storing at least one prior image from prior inspections. The system and method may further include registering the at least one current image with the at least one prior image, comparing the registered at least one current image with the at least one prior image to determine a transformation therebetween and updating the database with the at least one current image.
摘要:
A video-based fire detection system receives video data comprised of a plurality of individual frames, and determines based on the video data the ability of the system to detect the presence of fire. The system includes a video recognition system connectable to receive the video data and to calculate one or more background features associated with the video data. Based on the calculated background features, the video recognitions system assesses the ability of the video-based fire detection system to detect the presence of fire. The system includes one or more outputs operably connectable to communicate the results of the assessment made by the video recognitions system.
摘要:
A computer program product and method for performing position control on device members that have been identified as defective. The method may include providing a storage medium that stores data and programs used in processing images and a processing unit that processes the images, receiving, by the processing unit from an image capture device coupled to the processing unit, a set of images of a plurality of members inside of a device, detecting, by the processing unit, a defect in a first member of the plurality of members, and providing instructions to move the first member to an inspection position in the device. The device may be an engine and the members may be blades within the engine.
摘要:
A video enhancement system provides automatic enhancement to video data that includes the presence of obscurations. The system provides near-field enhancement of the video data by detecting the presence of near-field obscurations such as snowflakes/raindrops and determining whether the detected obscurations are located over background pixels or foreground objects. The detected obscuration pixels are filled-in depending on whether they are located over the background or foreground to create an enhanced image. The system may also provide local/global adaptive contrast enhancement to enhance video in the presence of far-field obscurations.
摘要:
A video content analysis (VCA) system generates an output regarding a detected condition that provides an indication of a confidence level regarding the detected condition. One example VCA system determines whether a first characteristic of a detected object in a field of vision of the video content analysis system satisfies a first criterion. If so, a first signal is generated under selected conditions. The VCA system also determines whether a second characteristic of the detected object satisfies a corresponding second criterion. If so, a second, different signal is generated if the first and second criteria are satisfied. The first and second signals indicate respective, different confidence levels that an event has occurred. A disclosed example includes a VCA as part of a security system.