1271 results about "Motion estimate" patented technology

Dense range data obtained at real-time rates is employed to estimate the pose of an articulated figure. In one approach, the range data is used in combination with a model of connected patches. Each patch is the planar convex hull of two circles, and a recursive procedure is carried out to determine an estimate of pose which most closely correlates to the range data. In another aspect of the invention, the dense range data is used in conjunction with image intensity information to improve pose tracking performance. The range information is used to determine the shape of an object, rather than assume a generic model or estimate structure from motion. In this aspect of the invention, a depth constraint equation, which is a counterpart to the classic brightness change constraint equation, is employed. Both constraints are used to jointly solve for motion estimates.

A method for head pose estimation may include receiving block motion vectors for a frame of video from a block motion estimator, selecting at least one block for analysis, determining an average motion vector for the at least one selected block, combining the average motion vectors over time (all past frames of video) to determine an accumulated average motion vector, estimating the orientation of a user's head in the video frame based on the accumulated average motion vector, and outputting at least one parameter indicative of the estimated orientation.

Pulse oximetry is improved through classification of plethysmographic signals by processing the plethysmographic signals using a neural network that receives input coefficients from multiple signal domains including, for example, spectral, bispectral, cepstral and Wavelet filtered signal domains. In one embodiment, a plethysmographic signal obtained from a patient is transformed (240) from a first domain to a plurality of different signal domains (242, 243, 244, 245) to obtain a corresponding plurality of transformed plethysmographic signals. A plurality of sets of coefficients derived from the transformed plethysmographic signals are selected and directed to an input layer (251) of a neural network (250). The plethysmographic signal is classified by an output layer (253) of the neural network (250) that is connected to the input layer (251) by one or more hidden layers (252).

The present invention provides a system (method and apparatus) for creating photorealistic 3D models of environments and / or objects from a plurality of stereo images obtained from a mobile stereo camera and optional monocular cameras. The cameras may be handheld, mounted on a mobile platform, manipulator or a positioning device. The system automatically detects and tracks features in image sequences and self-references the stereo camera in 6 degrees of freedom by matching the features to a database to track the camera motion, while building the database simultaneously. A motion estimate may be also provided from external sensors and fused with the motion computed from the images. Individual stereo pairs are processed to compute dense 3D data representing the scene and are transformed, using the estimated camera motion, into a common reference and fused together. The resulting 3D data is represented as point clouds, surfaces, or volumes. The present invention also provides a system (method and apparatus) for enhancing 3D models of environments or objects by registering information from additional sensors to improve model fidelity or to augment it with supplementary information by using a light pattern projector. The present invention also provides a system (method and apparatus) for generating photo-realistic 3D models of underground environments such as tunnels, mines, voids and caves, including automatic registration of the 3D models with pre-existing underground maps.

Method for computer-aided motion estimation in a plurality of temporally successive digital images. The method includes first partial motion estimating in a second digital image relative to a first digital image temporally preceding the second digital image; constructing a reference image structure from the first digital image and the second digital image based on the first partial motion estimation, the reference image structure containing at least features from the first digital image and / or the second digital image; second partial motion estimating in a third digital image, which temporally succeeds the second digital image, relative to the second digital image; third partial motion estimating with a comparison of features of the third digital image and of the features contained in the reference image structure; and determining motion in the third digital image relative to the first digital image based on the third partial motion estimation, the second partial motion estimation and the first partial motion estimation.

A device or method for video compression uses a technique in which changes in the image are encoded by motions of block of the image and signals indicating evolutions in the block. To determine the motions of the blocks of a each frame, a search is performed for a similar block of a previous frame based on points of the previous frame which are arranged in successive diamond shaped zones. The diamond shaped zones may be centred on the position of the block in previous frame, or one or more predicted motions of the block. The method terminates according to criteria defined using thresholds.

Motion estimation is the science of predicting the current frame in a video sequence from the past frame (or frames), by slicing it into rectangular blocks of pixels, and matching these to past such blocks. The displacement in the spatial position of the block in the current frame with respect to the past frame is called the motion vector. This method of temporally decorrelating the video sequence by finding the best matching blocks from past reference frames—motion estimation—makes up about 80% or more of the computation in a video encoder. That is, it is enormously expensive, and methods do so that are efficient are in high demand. Thus the field of motion estimation within video coding is rich in the breadth and diversity of approaches that have been put forward. Yet it is often the simplest methods that are the most effective. So it is in this case. While it is well-known that a full search over all possible positions within a fixed window is an optimal method in terms of performance, it is generally prohibitive in computation. In this patent disclosure, we define an efficient, new method of searching only a very sparse subset of possible displacement positions (or motion vectors) among all possible ones, to see if we can get a good enough match, and terminate early. This set of sparse subset of motion vectors is preselected, using a priori knowledge and extensive testing on video sequences, so that these “predictors” for the motion vector are essentially magic. The art of this method is the preselection of excellent sparse subsets of vectors, the smart thresholds for acceptance or rejection, and even in the order of the testing prior to decision.

A plurality of projection images are acquired over an angular range during the slow rotation of a C-arm gantry having a source and detector. Phase-specific reconstructions are generated from the plurality of projections, wherein each phase-specific reconstruction is generated generally from projections acquired at or near the respective phase. In one embodiment, a plurality of motion estimates are generated based upon the phase-specific reconstructions. One or more motion-corrected reconstructions may be generated using the respective motion estimates and projections. The motion-corrected reconstructions may be associated to form motion-corrected volume renderings.

A system and method for determining and correcting for imaging device motion during an exposure is provided. According to various embodiments of the present invention, multiple sets of image pixels are defined on an image sensor, where each set of pixels is at least partially contained in the output image area of the image sensor. Signals from each set of image pixels are read out once or more during an exposure, motion estimates are computed using signal readouts from one or more sets of image pixels, and signal readouts from one or more sets of the image pixels are processed to form the final output image.

Embodiments include a motion estimation method performed in a parallel processing system that determines a list of several candidate motion vectors for a macroblock of a video image and retains them through multiple computation passes. All candidate motion vectors are used as potential neighboring predictors, so that the best combination of differential vectors rises to the top of the candidate list. Numerous combinations of differential motion vectors are considered during the process that compares motion vectors among up to eight neighboring macroblocks, instead of simply between pairs of macroblocks. The motion estimation system is configured to use a large number of compute engines, such as on a highly parallel GPU platform. This is achieved by having no dependencies between macroblocks except one per pass. This allows the number of calculations per pass to be very large.

An interlace-to-progressive scan conversion system comprises: a spatial line averaging prefilter; a motion estimator; a three-stage adaptive recursive filter. The motion estimator comprises: a 3-D recursive search sub-component having a bilinear interpolator; a motion correction sub-component having an error-function including penalties related to the difference between a given candidate vector and a plurality of neighboring vectors; a block erosion sub-component. The motion estimator assumes that motion is constant between fields. The three-stage adaptive recursive filter comprises: a first stage that selects between using static pixels data and moving pixels data from a next field; a second stage that selects a more valid set of data between motion compensated data from a previous field and the pixels selected by the first stage; a third stage that combines an intra-field interpolation with the more valid set of data selected by the second stage.

A genus of motion estimation processes is disclosed which is characterized by the following characteristics which all species in the genus will share 1) a process within this genus does not perform the motion estimation separately for each of the partitions and subpartitions defined in the H.264 standard; 2) a process within the genus computes for each motion vector in the search region the partial costs for all macroblock partitions and sub-partitions, compares them to the best partial costs found so far, and for partitions and sub-partitions having lower costs updates the corresponding best partial costs and records the current motion vectors as the one realizing them. 3) a process within the genus after finishing scanning the motion vectors in the search region, computes from the best partial costs the total costs for all possible macroblock partitioning modes and selects the one with the lowest total cost as the best macroblock partitioning mode, with the best motion vectors corresponding to each of the selected macroblock partitions and sub-partitions.

A system for estimating orientation of a target based on real-time video data uses depth data included in the video to determine the estimated orientation. The system includes a time-of-flight camera capable of depth sensing within a depth window. The camera outputs hybrid image data (color and depth). Segmentation is performed to determine the location of the target within the image. Tracking is used to follow the target location from frame to frame. During a training mode, a target-specific training image set is collected with a corresponding orientation associated with each frame. During an estimation mode, a classifier compares new images with the stored training set to determine an estimated orientation. A motion estimation approach uses an accumulated rotation / translation parameter calculation based on optical flow and depth constrains. The parameters are reset to a reference value each time the image corresponds to a dominant orientation.

Dense range data obtained at real-time rates is employed to estimate the pose of an articulated figure. In one approach, the range data is used in combination with a model of connected patches. Each patch is the planar convex hull of two circles, and a recursive procedure is carried out to determine an estimate of pose which most closely correlates to the range data. In another aspect of the invention, the dense range data is used in conjunction with image intensity information to improve pose tracking performance. The range information is used to determine the shape of an object, rather than assume a generic model or estimate structure from motion. In this aspect of the invention, a depth constraint equation, which is a counterpart to the classic brightness change constraint equation, is employed. Both constraints are used to jointly solve for motion estimates.

In accordance with an embodiment of the present invention, an efficient and accurate system and method are described for detecting a fingerprint from swipe sensor system. A swipe sensor module is coupled to a microprocessor module. The swipe sensor module collects fingerprint image data as a plurality of frames and passes the frames of data to the microprocessor module. The microprocessor module assembles the frames of data into a complete image of the fingerprint that is cropped and processed to remove noise and artifacts. The cropped image is used to generate a template in a first instance and to compare an extracted portion of the cropped image to existing templates in a second instance.

A method for tracking one or multiple objects from an input video sequence allows a user to select one or more regions that contain the object(s) of interest in the first and the last frame of their choice. An initialization component selects the current and the search frame and divides the selected region into equal sized macroblocks. An edge detection component computes the gradient of the current frame for each macroblock and a threshold component decides then which of the macroblocks contain sufficient information for tracking the desired object. A motion estimation component computes for each macroblock in the current frame its position in the search frame. The motion estimation component utilizes a search component that executes a novel search algorithm to find the best match. The mean absolute difference between two macroblocks is used as the matching criterion. The motion estimation component returns the estimated displacement vector for each block. An output component collects the motion vectors of all the predicted blocks and calculates the new position of the object in the next frame.

A motion estimating method of H.264 / AVC coder divides integer motion estimation to be rough layer and fine layer being estimated separately in sequence then inputting parallel interpolation of half pixel and fine layer motion estimation into reference frame data and integer pixel data of rough layer motion vector separately and carrying out half pixel motion estimation as well as quarter pixel motion estimation for calculating out final minimum residual error and its corresponding final motion vector by utilizing motion vector corresponding to minimum residual error under optimum prediction mode obtained from rough layer motion estimation.

Provided are a method and device for motion estimation and compensation to be performed on a panorama image. The motion estimation and compensation are performed on a panorama image with a 360° omni-directional view based on that the spatial relation between left and right borders of the panorama image is very high. Accordingly, it is possible to improve image quality through effective and precise estimation and compensation for the motion of a panorama image. In particular, it is possible to improve the image quality at the right and left edges of the panorama image.

An improved treadmill control system which adjusts the speed of a moving tread belt to follow user motions. Equipment includes a tread base supporting a moving tread belt upon which a user can run or walk, a motor assembly and motor driver to move the tread belt, a plurality of foot sensors, a tread belt motion sensor, a measurement system to estimate user motion based on foot and tread belt sensor signals, and a motor controller to adjust motor assembly speed based on estimates of user motion. The system is capable of making improved user motion estimates and of using them to provide improved belt speed control. In one embodiment, user position, speed, and acceleration are estimated at each user footfall while estimates are continually revised between footfalls. In one embodiment, foot sensors are capacitive proximity sensors which are effective, fully concealable, and economical.

Method, systems and software are proposed for obtaining for blocks of a first image similar blocks of a second image (the “reference image”). The blocks of the first image are processed sequentially, for each block trying out a number of candidate locations in the second image and evaluating a cost function for each. Each candidate location in the second image is displaced by a respective motion vector from the block of the first image. In a first aspect of the invention the cost function is a function of a predicted motion vector for future blocks of the first image (i.e. blocks of the first image which have not yet been processed). In a second aspect of the invention the motion vectors are given by location values which are not all whole pixel spacings, halves of the pixel spacing, or quarters of the pixel spacing.

The present invention provides a method to compress digital moving pictures or video signals based on an overcomplete basis transform using a modified Matching Pursuit algorithm. More particularly, this invention focuses on the efficient coding of the motion residual image, which is generated by the process of motion estimation and compensation. A residual energy segmentation algorithm (RESA) can be used to obtain an initial estimate of the shape and position of high-energy regions in the residual image. A progressive elimination algorithm (PEA) can be used to reduce the number of matching evaluations in the matching pursuits process. RESA and PEA can speed up the encoder by many times for finding the matched basis from the pre-specified overcomplete basis dictionary. Three parameters of the matched pattern form an atom, which defines the index into the dictionary and the position of the selected basis, as well as the inner product between the chosen basis pattern and the residual signal. The present invention provides a new atom position coding method using quad tree like techniques and a new atom modulus quantization scheme. A simple and efficient adaptive mechanism is provided for the quantization and position coding design to allow a system according to the present invention to operate properly in low, medium and high bit rate situations. These new algorithm components can result in a faster encoding process and improved compression performance over previous matching pursuit based video coders.

An imaging device, comprising: an image pickup device which images a subject; and a motion estimating device which estimates a motion of the subject and a motion of the imaging device based on a dynamic image or a plurality of still images which are imaged to include the subject prior to actual imaging of a still image of the subject by the image pickup device.

A method for estimating the motion between a first image and a second image in a system for creating a panoramic image from a plurality of images taken by a camera, the method comprising the steps of: downsampling a first image in a first direction and in a second direction; downsampling a second image in a first direction and in a second direction; filtering the first and the second image so as to filter out any global illumination changes between the first image and the second image; calculating a first displacement along the first direction between the first downsampled image along the first direction and the second downsampled image along the first direction; and calculating a second displacement along the second direction between the first downsampled image along the second direction and the second downsampled image along the second direction. In an alternate embodiment, a device and computer readable medium corresponding to the above method is described.

A motion estimation algorithm finds the best match for a given block or macroblock so that the resulting error signal has very low energy level which is computed, for e.g., by the SAD method. The motion estimation algorithm also provides for an optional sub-pixel level estimation and an inter4v search, and allows for restricting the number of searches for a Frame-frame ME (motion estimation) using Top-Top and Bottom-Bottom field MEs. The algorithm provides for a selective early exit and enables selecting a suitable search area with N candidate points (4 to 8) for starting the search. The search is conducted progressively till a minimum error signal (low energy level signal) is reached. The candidate points for search may be in a diamond shaped configuration, and there may be a plurality of successive diamond configurations, the number of which is configurable. The invention has application in MPEG-4 and H.264 standards.

Motion vectors for encoding a predicted frame relative to a reference frame are determined from a phase correlation of corresponding regions of the predicted frame and reference frame. Peaks in the phase correlation are identified, and the location of the peaks are used as candidate motion vectors. From this limited set of candidate motion vectors, the best motion vectors for predicting blocks within each region can be readily identified.

A method, a circuit arrangement and a display device which suppresses motion blur due to motion integration carried out along a motion trajectory on the image data, which occurs in matrix type displays in which a motion trajectory is integrated by the viewer and / or the display, includes an inverse integration filtering of the video signal. To avoid de-blurring in image regions where motion cannot be detected, or image is low and to avoid noise modulation, motion estimation as well as further image characteristic are used.

A video coding method for surveillance videos allowing some regions of the scene to be encoded in an almost lossless manner. Such Regions of Interest (RoI) can be determined a priori or they can be automatically determined in real-time by an intelligent system. The user can set high priority in such regions a priori or the intelligent video analysis algorithm can automatically assign some windows a higher priority compared to the rest of the video. In a preferred embodiment, this can be achieved by canceling the motion estimation and compensation operations, and then decreasing the size of the quantization levels during the encoding process in the RoI. The present inventions can produce MPEG compatible bit-streams without sending any side information specifying the RoI.

A method to generate super-resolution images using a sequence of low resolution images is disclosed. The method includes generating an estimated high resolution image, motion estimating between the estimated high resolution image and comparison images from the sequence of low resolution images, motion-compensated back projecting, and motion-free back projecting that results in a super resolved image. A corresponding system for generating super-resolution images includes a high resolution image estimation module, a motion estimating module, a motion-compensated back projection module, a motion-free back projection module, an input interface, and an output interface.