Abstract: A system and method for automatic scale selection in real-time image and video processing and computer vision applications. In one aspect, a non-parametric variable bandwidth mean shift technique, which is based on adaptive estimation of a normalized density gradient, is used for detecting one or more modes in the underlying data and clustering the underlying data. In another aspect, a data-driven bandwidth (or scale) selection technique is provided for the variable bandwidth mean shift method, which estimates for each data point the covariance matrix that is the most stable across a plurality of scales. The methods can be used for detecting modes and clustering data for various types of data such as image data, video data speech data, handwriting data, etc.

Abstract: A system and method for automatic scale selection in real-time image and video processing and computer vision applications. In one aspect, a non-parametric variable bandwidth mean shift technique, which is based on adaptive estimation of a normalized density gradient, is used for detecting one or more modes in the underlying data and clustering the underlying data. In another aspect, a data-driven bandwidth (or scale) selection technique is provided for the variable bandwidth mean shift method, which estimates for each data point the covariance matrix that is the most stable across a plurality of scales. The methods can be used for detecting modes and clustering data for various types of data such as image data, video data speech data, handwriting data, etc.

Abstract: A system and method for automatic scale selection in real-time image and video processing and computer vision applications. In one aspect, a non-parametric variable bandwidth mean shift technique, which is based on adaptive estimation of a normalized density gradient, is used for detecting one or more modes in the underlying data and clustering the underlying data. In another aspect, a data-driven bandwidth (or scale) selection technique is provided for the variable bandwidth mean shift method, which estimates for each data point the covariance matrix that is the most stable across a plurality of scales. The methods can be used for detecting modes and clustering data for various types of data such as image data, video data speech data, handwriting data, etc.

Abstract: A system and method for object detection with sudden illumination changes provides a system including a camera model unit for computing a difference map between a first input image and a second input image, an illumination model unit in signal communication with the camera model unit for comparing the sign of the difference between two pixels in a first input image against the sign of the difference between two pixels in a second input image, and a CPU in signal communication with the camera model unit and the illumination model unit for detecting a changed object in accordance with said comparison when said sign of the difference is changed; and a corresponding method including providing a first image having a number of pixels, receiving a second image subsequent to the first image, calculating a difference map between the first and second images, selecting a pair of pixel locations from the first or second images in accordance with the difference map, statistically modeling a camera noise level for the firs

Abstract: The present invention relates to a method for visually detecting and tracking an object through a space. The method chooses modules for a restricting a search function within the space to regions with a high probability of significant change, the search function operating on images supplied by a camera. The method also derives statistical models for errors, including quantifying an indexing step performed by an indexing module, and tuning system parameters. Further the method applies a likelihood model for candidate hypothesis evaluation and object parameters estimation for locating the object.

Abstract: A method for detecting and recognizing at least one traffic sign is disclosed. A video sequence having a plurality of image frames is received. One or more filters are used to measure features in at least one image frame indicative of an object of interest. The measured features are combined and aggregated into a score indicating possible presence of an object. The scores are fused over multiple image frames for a robust detection. If a score indicates possible presence of an object in an area of the image frame, the area is aligned with a model. A determination is then made as to whether the area indicates a traffic sign. If the area indicates a traffic sign, the area is classified into a particular type of traffic sign. The present invention is also directed to training a system to detect and recognize traffic signs.

Abstract: A system and method for vehicle detection and tracking in tunnels is provided. The method comprises the steps of capturing a plurality of image frames viewing at least one traffic lane; extracting at least one feature from the plurality of image frames; detecting at least one object indicative of a vehicle from the extracted feature; and tracking the detected vehicle over time to determine the detected vehicle's velocity. The system comprising at least one image capture device for capturing a plurality of image frames viewing at least one traffic lane; and a processor adapted for extracting at least one feature from the plurality of image frames, detecting at least one object indicative of a vehicle from the extracted feature, and tracking the detected vehicle over time to determine the detected vehicle's velocity.

Abstract: A method for dynamic scene modeling and change detection applicable to motion analysis utilizes optical flow for capturing and modeling the dynamics of the scene. Uncertainties in the measurements are evaluated and utilized in order to develop a robust representation of the scene in a higher dimensional space. In another embodiment, a dynamical model of the scene is developed that utilizes multiple past frames to predict the next frame. Incremental methods for updating the model are developed and, towards detection of events, a new measure is introduced that is based on a state-driven comparison between the prediction and the actual observation.

Abstract: An imaging system (110) for imaging a scene with a detector array (104) having an array of imaging elements is provided. The imaging system (110) includes an image estimation module (202) for generating a plurality estimates of uncorrupted images based upon a plurality of noisy images generated by the detector array (104). The imaging system (110) further includes a parameter determination module (204) for determining one or more nonuniformity correction parameters based upon the estimates of uncorrupted images.

Abstract: A system and corresponding method for image acquisition are provided, the system including a processor, an imaging adapter in signal communication with the processor for receiving image data from each of a static imaging device and a dynamic imaging device, and a homography unit in signal communication with the processor for computing a planar homography between the static and dynamic image data; and the method including receiving an image from a static imaging device, receiving an image from a dynamic imaging device, and registering the dynamic image to the static image using planar homography.

Abstract: A system and method for illumination invariant change detection are provided, the system including a processor, an energy ranking unit in signal communication with the processor for extracting block coefficients for the first and second images and computing an energy difference responsive to the coefficients for a frequency energy between the first and second images, and a change detection unit in signal communication with the processor for analyzing the energy difference and detecting a scene change if the energy difference is indicative of change; and the method including receiving first and second images, extracting block coefficients corresponding to frequency energies for the first and second images, computing an energy difference for at least one of the frequency energies between the first and second images, analyzing the at least one energy difference, and detecting a scene change if the energy difference is indicative of change.

Abstract: A method for detecting events in a video sequence includes providing a video sequence, sampling the video sequence at regular intervals to form a series of snapshots of the sequence, measuring a similarity of each snapshot, measuring a similarity change between successive pairs of snapshots, wherein if a similarity change magnitude is greater than a predetermined threshold, a change event has been detected, verifying the change event to exclude a false positive, and completing the processing of the snapshot incorporating the verified change event.

Abstract: A system and method for object detection are provided where the system includes a component detection unit for detecting components in an image, a component fusion unit in signal communication with the component detection unit for fusing the components into an object, and a CPU in signal communication with the detection and fusion units for comparing the fused components with a statistical model; and the method includes receiving observation data for a plurality of training images, forming at least one statistical model from the plurality of training images, receiving an input image having a plurality of pixels, detecting a plurality of components in the input image, determining a fusion of the detected components, comparing the fusion with the statistical model, and detecting an object in accordance with the comparison.

Abstract: A client-server system and method that enables efficient, low bit rate transmission of image data over a network from an image server (e.g., active cameras) to a client for, e.g., distributed surveillance. A detection and tracking module detects a human presence within an observation area and provides 2-dimensional face coordinates and its estimated scale to a video transmission module. The captured video is then efficiently encoded in log-polar coordinates using an adaptive log-polar mapping, with a foveation point centered on the face. A fovea region, which covers the target object (face), is uniformly sampled and transmitted at full resolution to the client. The periphery region (background) is sampled according to a log-polar grid.

Abstract: A system and method for providing decision support to a physician during a medical examination is disclosed. Data is received from a sensor representing a particular medical measurement. The received data includes image data. The received data and context data is analyzed with respect to one or more sets of training models. Probability values for the particular medical measurement and other measurements to be taken are derived based on the analysis and based on identified classes. The received image data is compared with training images. Distance values are determined between the received image data and the training images, and the training images are associated with the identified classes. Absolute value feature sensitivity scores are derived for the particular medical measurement and other measurements to be taken based on the analysis. The probability values, distance values and absolute value feature sensitivity scores are outputted to the user.

Abstract: A system and method for vehicle detection and tracking in tunnels is provided. The method comprises the steps of capturing a plurality of image frames viewing at least one traffic lane; extracting at least one feature from the plurality of image frames; detecting at least one object indicative of a vehicle from the extracted feature; and tracking the detected vehicle over time to determine the detected vehicle's velocity. The system comprising at least one image capture device for capturing a plurality of image frames viewing at least one traffic lane; and a processor adapted for extracting at least one feature from the plurality of image frames, detecting at least one object indicative of a vehicle from the extracted feature, and tracking the detected vehicle over time to determine the detected vehicle's velocity.

Abstract: A method of tracking an object comprises providing a plurality of cameras, determining an image from each camera, and determining a common plane in the images. The method further comprises determining a parallax for scene points across the images, incorporating the parallax as a feature in a background model, and estimating a change in the scene using the background model. A camera can be a pan-tilt-zoom camera. A camera can be uncalibrated.

Abstract: The present invention comprises using error propagation for building feature spaces with variable uncertainty and using variable-bandwidth mean shift for the analysis of such spaces, to provide peak detection and space partitioning. The invention applies these techniques to construct and analyze Hough spaces for line and geometrical shape detection, as well as to detect objects that are represented by peaks in the Hough space. This invention can be further used for background modeling by taking into account the uncertainty of the transformed image color and uncertainty of the motion flow. Furthermore, the invention can be used to segment video data in invariant spaces, by propagating the uncertainty from the original space and using the variable-bandwidth mean shift to detect peaks. The invention can be used in a variety of applications such as medical, surveillance, monitoring, automotive, augmented reality, and inspection.

Abstract: This invention relates to the integration of visual information, anatomic constraints, prior shape knowledge, and level set representations for the segmentation of medical images. An embodiment according to the present invention comprises a level set variational framework that uses a bi-directional boundary flow, an intensity-based regional component that maximizes the a posteriori segmentation probability, a physiology-based module that constrains the solution space and a term that accounts for shape-driven consistency. All modules are expressed in an energetic form and the resulting objective function is optimized using a gradient descent method.

Abstract: A system and method for non-rigid image registration using distance functions includes portions for receiving a source shape into an image space, integrating a global linear registration model with local deformations to assess the source shape, optimizing a functional defined on a parameter or feature space where the functional quantifies the similarity between the source shape and a target shape in terms of distance functions, creating an augmented registration space including a plurality of target shape clones coherently positioned in the image space, tracking moving interfaces between the source shape and the target shapes and clones through a level set method, and registering the source shape by seeking mutual correspondences between the source shape, the target shape and the target shape clones.