3139 results about "Euclidean distance" patented technology

A system and method for rescoring the N-best hypotheses from an automatic speech recognition system by comparing an original speech waveform to synthetic speech waveforms that are generated for each text sequence of the N-best hypotheses. A distance is calculated from the original speech waveform to each of the synthesized waveforms, and the text associated with the synthesized waveform that is determined to be closest to the original waveform is selected as the final hypothesis. The original waveform and each synthesized waveform are aligned to a corresponding text sequence on a phoneme level. The mean of the feature vectors which align to each phoneme is computed for the original waveform as well as for each of the synthesized hypotheses. The distance of a synthesized hypothesis to the original speech signal is then computed as the sum over all phonemes in the hypothesis of the Euclidean distance between the means of the feature vectors of the frames aligning to that phoneme for the original and the synthesized signals. The text of the hypothesis which is closest under the above metric to the original waveform is chosen as the final system output.

A method and apparatus for animating a synthetic body part. The 3D-animation system and method use a database of basic postures. In a first step, for each frame, a linear combination of the basic postures from a database of basic postures is obtained by minimizing the Euclidean distance between the displacement of critical points. The displacement information is supplied externally, and typically can be obtained by observing the displacement of physical markers placed on a moving physical body part in the real world. For instance, the synthetic body part may be an expression of a human face and the displacement data are obtained by observing physical markers placed on the face of an actor. The linear combination of the postures in the database of postures is then used to construct the desired posture. Postures are constructed for each time frame and are then displayed consecutively to provide animation. A computer readable storage medium containing a program element to direct a processor of a computer to implement the animation process is also provided.

The invention discloses a three-dimensional reconstruction method based on coding structured light, comprising the following steps: 1) projecting structured light to an object to be measured, and capturing an image modulated by the object to be measured by a camera; 2) matching an optical template, comprising: (2.1) positioning the optical strip boundary, scanning along each row of the image, determining a pixel point with strong gray variation as a candidate marginal point, and searching a local domain; and (2.2) matching the optical strip: adopting a color cluster method to build a color matching proper vector, comparing image color with a projected color, and defining Euclidean distance between the color proper vector and the cluster center to distribute the colors of red, green, blue and white of the candidate optical strip; and 3) using a calibrated system parameter for three-dimensional reconstruction of the object to be measured, determining the relation between a space point coordinate and the image coordinate point thereof by the calibrated conversion matrix parameter; and restoring three-dimensional spatial coordinate from the image coordinate of a feature point. The invention can simplify calculation process and has high matching precision and high reconstruction precision.

The invention is a computer implemented technique for id entifying anomalous data in a data set. The method uses cascaded k-Means clustering and the ID3 decision tree learning methods to characterize a training data set having data points with known characterization. The k-Means clustering method first partitions the training instances into k clusters using Euclidean distance similarity. On each training cluster, representing a density region of normal or anomaly instances, the invention builds an ID3 decision tree. The decision tree on each cluster refines the decision boundaries by learning the sub-groups within the cluster. A test data point is then subjected to the clustering and decision trees constructed form the training instances. To obtain a final decision on classification, the decisions of the k-Means and ID3 methods are combined using rules: (1) the Nearest-neighbor rule, and (2) the Nearest-consensus rule.

The invention provides a remote sensing image registration method of a multi-source sensor, relating to an image processing technology. The remote sensing image registration method comprises the following steps of: respectively carrying out scale-invariant feature transform (SIFT) on a reference image and a registration image, extracting feature points, calculating the nearest Euclidean distances and the nearer Euclidean distances of the feature points in the image to be registered and the reference image, and screening an optimal matching point pair according to a ratio; rejecting error registration points through a random consistency sampling algorithm, and screening an original registration point pair; calculating distribution quality parameters of feature point pairs and selecting effective control point parts with uniform distribution according to a feature point weight coefficient; searching an optimal registration point in control points of the image to be registered according to a mutual information assimilation judging criteria, thus obtaining an optimal registration point pair of the control points; and acquiring a geometric deformation parameter of the image to be registered by polynomial parameter transformation, thus realizing the accurate registration of the image to be registered and the reference image. The remote sensing image registration method provided by the invention has the advantages of high calculation speed and high registration precision, and can meet the registration requirements of a multi-sensor, multi-temporal and multi-view remote sensing image.

A method for detecting moving objects includes: (a) capturing and establishing a background image; (b) capturing at least one current image; (c) transforming the background image and the current image from an RGB color format into an HSI color format; (d) subtracting the background image from the current image according to a background subtraction rule for generating at least one moving object; (e) performing a vertical scanning and a horizontal scanning on the moving object for generating a minimum bounding box of the moving object; (f) calculating a characteristic datum of the moving object according to the minimum bounding box; (g) tracking the moving object according to the characteristic datum with a Euclidean distance rule; (h) classifying the moving object according to the characteristic datum, the tracking result generated by step (g) and a minimum distance classifier.

The invention discloses a pedestrian re-identification method based on multi-attribute and multi-strategy fusion learning. The method of the invention includes the steps of in an offline training phase, firstly selecting pedestrian attributes which are easy to be judged and have a sufficient distinguishing degree, training a pedestrian attribute identifier on an attribute data set, then labeling attribute tags for a pedestrian re-identification data set by using the attribute identifier, and next, by combining the attributes and pedestrian identity tags, training a pedestrian re-identification model by using a strategy fused with pedestrian classification and novel constraint comparison verification; and in an online query phase, extracting features of a query image and images in a database by using the pedestrian re-identification model, and calculating the Euclidean distance between the feature of the query image and the feature of each image in the database to obtain the image with the shortest distance, which is considered as the result of pedestrian re-identification. In terms of performance, the features in the invention are distinguishable and high accuracy is obtained; and in terms of efficiency, the method of the invention can quickly search for the pedestrian indicated by the query image from the pedestrian image database.

A system for transmitting data over a MIMO channel has a transmitter and a receiver. In the transmitter, the input data is encoded over at least two pipes by a concatenation of at least two constituent signal-space encoders. Each constituent encoder is used to generate, in response to the input data, a sequence of symbols from a channel alphabet having at least one dimension. Each symbol of the channel alphabet includes at least one complex symbol having real and imaginary coordinates. The transmitter interleaves the coordinates of the sequence of channel alphabet symbols, and transmits (from at least two transmit antennas) the interleaved coordinates. Preferably, each constituent encoder maximizes a minimum coordinate-wise Hamming distance between members of all valid pairs of symbol sequences, maximizes a minimum Euclidean distance between members of all valid pairs of different codewords, and obeys an equal eigenvalue criterion.

A communication device for simultaneously transmitting information on multiple sub-channels encodes information for each of the multiple sub-channels with a coding scheme to produce channel encoded information. A mask vector derived from a redundancy in the coding scheme, encodes the channel encoded information to transform the channel encoded information into codewords having pairwise Euclidean distance properties identical to those of the channel encoded information, Modulation of the sub-channels in accordance with the codewords in a modulator then produces a composite signal envelope having a peak-to-mean error power ratio (PMEPR) reduced relative to a PMEPR for correspondingly modulated channel encoded information.

The invention discloses a cold-chain logistic stowage intelligent recommendation method based on spectral clustering. Scores of users for a stowage line are conveyed through a cold chain for cold-chain logistic stowage intelligent recommending, a score matrix is built, the Euclidean distance is used for calculating the user similarity, a degree matrix is used for calculating a Laplacian matrix, feature vectors are obtained by calculating feature values of the orderly Laplacian matrix, a K-means algorithm is used for clustering the feature values to obtain a user group with the similar interesting stowage line, and a stowage line is recommended inside the user group with the similar interesting stowage line, so that cold-chain logistic stowage intelligent recommending is achieved, the cold-chain logistic vehicle non-load ratio is lowered, and the profit rate of cold-chain logistic transport vehicles is increased.

A method of evaluating the quality of a read signal from the viewpoint of the detection margin of a Viterbi decoder in the PRML method in which a target signal level varies depending on the read signal, and an optical disc apparatus implementing the method. A method of evaluating the quality of a signal pattern comprising a combination of minimum run lengths from the viewpoint of edge shift, and an optical disc apparatus implementing the method. From the target signal level that varies depending on the read signal, a target signal is generated based on a decoding result, and an error target signal is generated in which the decoding result is edge-shifted. The signal quality is evaluated by calculating a Euclidean distance between these signals and the read signal. A virtual state that is not included in the Viterbi decoder and that is less than the minimum run length is defined, and a target signal level for the virtual state is generated using a target signal level table inside the Viterbi decoder, based on the concept of convolution. In this way, the signal quality can be evaluated by the same method as mentioned above even in cases where the pattern of a combination of the minimum run lengths has edge-shifted.

The present invention provides an indoor mobile robot vision SLAM method based on the Kinect. The method comprises the following steps: S1: acquiring color RGB data and Depth data of an indoor environment by using a Kinect camera; S2: performing feature detection of RGB data and implementing rapid and effective matching between adjacent images; S3: combining inner parameters of a Kinect camera after calibration and pixel point depth values to convert a 2D image point into a 3D space point, and establishing a corresponding relationship of 3D point cloud; S4: using the RANSAC algorithm to eliminate external points of the point cloud to complete point cloud rough matching; S5: employing an ICP algorithm with a double limit of an Euclidean distance and an angle threshold to complete fine matching of the point cloud; and S6: introducing a weight in a key frame selection, and employing the g2o algorithm to optimize a robot posture, and finally obtaining a robot moving trajectory, and generating a 3D point cloud map. The indoor mobile robot vision SLAM method based on the Kinect can solve the problems that a point cloud registration portion in a vision SLAM system is liable to local optimum and is large the matching error, and therefore the registration accuracy of the point cloud is improved.

The invention discloses a method for optimizing WLAN (Wireless Local Area Network) indoor ANN (Artificial Neural Network) positioning based on FCM (fuzzy C-means) and least-squares curve surface fitting methods, relating to an indoor positioning method used for indoor positioning and aiming to solve generalization capability reduction of an ANN system caused by singular reference points existing in a training sample space. The method comprises the following steps of carrying out clustering on pre-labeled reference points based on the FCM method to confirm corresponding clustering centers and membership degree of different reference points to clustering centers thereof; obtaining the space position of the singular reference points in a target positioning area on the basis of carrying out quantitative processing and similarity calculation on the membership degree of the reference points; updating positioning fingerprint database by utilizing the least-squares curve surface fitting method to reject abrupt change points in an intensity distribution chart; estimating the cluster of a terminal on the basis of calculating the Euclidean distance between signal intensity samples collected online and different clustering centers; and finally accurately estimating the terminal by utilizing corresponding ANN subsystems.

The invention provides an unmanned aerial vehicle trajectory planning method based on EB-RRT. The method comprises the steps that grid partitioning is carried out on a map environment; a node xnearst nearest to a random point in existing nodes is found; an insertion point xnew is calculated according to the step length; if the sum of the distance between a root node to xnew and the Euclidean distance between xnew and the end is not greater than the length of the current optimal path, whether the xnew point is in an obstacle is detected; if not, the surrounding environment information of xnearst is collected, and a new insertion point xnew is randomly sampled in the surrounding free area; xnew is inserted into a tree; the nearby node set of xnew is traversed and found in the corresponding grid, and the path of the nearby nodes is optimized; connection detection is carried out on two trees into which the xnew point is inserted; if not, two trees are exchanged, and random points continue to be sampled; if so, a feasible path is found, and downsampling is carried out; and a Bessel cubic interpolation algorithm is used to optimize the new path. The unmanned aerial vehicle trajectory planning method provided by the invention has the advantages of high convergence speed, good flexibility, high running efficiency and good practicability.

The present invention relates to a system and methodology to facilitate database processing in accordance with a plurality of various applications. In one aspect, a large database of objects is processed, wherein the objects can be represented as points in a vector space, and two or more objects are deemed ‘close’ if a Euclidean distance between the points is small. This can apply for substantially any type of object, provided a suitable distance measure can be defined. In another aspect, a ‘test’ object having a vector x, is processed to determine if there exists an object y in the database such that the distance between x and y falls below a threshold t. If several objects in the database satisfy this criteria, a list of objects can be returned, together with their corresponding distances. If no objects were to satisfy the criterion, an indication of this condition can also be provided, but in addition, the condition or information relating to the condition can be provided.

The invention discloses a multi-touch trajectory tracking method. The method comprises the following steps of: A, calculating initial states of each touch point and acquiring the data of each touch point; B, predicting the possible position of each touch point in the next frame by a semi-Kalman filtering method; C, in a given threshold range, searching the touch point which is in the nearest Euclidean distance of a predicted position in a current frame; if the touch point is searched, performing correlation; and otherwise, taking a predicted value as the position of the touch point in the current frame; D, judging whether a new touch point is added into the current frame; a new touch point is detected, performing the step A on the new touch point, and performing a step E on other touch points; and E, judging whether a trajectory is ended in the current frame. In the method, the initial states and the ending states of the touch points are judged by a three-frame iteration method; the probability of misjudging a noise into the touch point is avoided to a certain degree; the influence of the noise on a correct trajectory is improved; and particularly the correctness of the judgment of the new touch point and the ending trajectory is improved.

The invention provides a method and a system for detecting DNS (domain name system) traffic abnormality. The method includes: extracting corresponding characteristic values for DNS traffic data to be processed, giving different weights to each characteristic, and detecting an abnormality cluster marked in a training set by the aid of the W-Kmeans algorithm and the additional Euclidean distance threshold Dthreshold so that new unknown characteristic abnormality can be discovered. The method and the system have the advantages that the algorithm is high in convergence speed and small in calculation, new samples to be detected only need to be compared with a processed training clustering center, calculation of a great quantity of original training data is not needed, the method and the system are low in deployment cost, strong in generalization ability and capable of discovering DNS traffic abnormality rapidly and effectively, and the system is particularly suitable for being deployed on a large DNS server.

The invention relates to an image mosaic method based on neighborhood Zernike pseudo-matrix of characteristic points, which leads image statistical information (Zernike pseudo-matrix) to be interrelated with the characteristic points of an image by using the following steps: firstly, extracting interest points of an inference image and an input image by utilizing a Harris angle detector, and taking rectangular neighborhoods using the interest points as a center as a local characteristic region with characteristic matching; secondly, calculating the Zernike pseudo-matrix to the rectangular characteristic regions to be used as descriptors of the characteristic region, and realizing the matching of the characteristic points through comparing the Euclidean distance of the characteristic vector of the descriptor of each characteristic region. The matching can have less wrong matching points which are eliminated through a RANSAC (RANdom Sample Consensus) algorithm, and the right matching relation can be calculated to realize the image mosaic. The invention can effectively realize the image registration and mosaic with geometric transformation relations of translation, rotation, small scale zooming, and the like, and can be used for image treatment and image synthesis of fields such as communication, multimedia technology, and the like.

The embodiment of the invention provides an integration target attribute identification and precise retrieval method based on depth measurement learning. The method comprises the following steps: obtaining a query picture and a picture database; according to a pre-trained deep neural network model, independently obtaining the feature vector of the query picture and the feature vectors of all pictures in the picture database, wherein the feature vectors comprise the category features of an individual target and the identity features of the individual target in the pictures; according to the feature vectors, independently calculating an Euclidean distance between the query picture and each picture in the picture database in an Euclidean space; and according to the Euclidean distances between the query picture and all pictures in the picture database, selecting a picture which exhibits highest similarity with the query picture in the picture database. The method can improve the accuracy of image retrieval.