摘要:
A visual media de-identification system is described. The system includes an image merger and a de-identifying engine. The image merger is configured to merge a sequence of images from a set of visual media data into an averaged image. The de-identifying engine is configured to: bound portions of the averaged image that are determined to be relatively fixed, wherein each bounded portion is identified by a corresponding position in the averaged image; generate a template comprising the bounded portions and the corresponding position for each bounded portion in the averaged image; and de-identify the sequence of images by obfuscating content in the bounded portions.
摘要:
A method for de-identification of visual media data, including: merging a sequence of images from a set of visual media data into an averaged image; bounding portions of the averaged image that are determined to be relatively fixed, wherein each bounded portion is identified by a corresponding position in the averaged image; generating a template comprising the bounded portions and the corresponding position for each bounded portion in the averaged image; and de-identifying the sequence of images by obfuscating content in the bounded portions.
摘要:
A method for de-identification of visual media data, including: merging a sequence of images from a set of visual media data into an averaged image; bounding portions of the averaged image that are determined to be relatively fixed, wherein each bounded portion is identified by a corresponding position in the averaged image; generating a template comprising the bounded portions and the corresponding position for each bounded portion in the averaged image; and de-identifying the sequence of images by obfuscating content in the bounded portions.
摘要:
A visual media de-identification system is described. The system includes an image merger and a de-identifying engine. The image merger is configured to merge a sequence of images from a set of visual media data into an averaged image. The de-identifying engine is configured to: bound portions of the averaged image that are determined to be relatively fixed, wherein each bounded portion is identified by a corresponding position in the averaged image; generate a template comprising the bounded portions and the corresponding position for each bounded portion in the averaged image; and de-identify the sequence of images by obfuscating content in the bounded portions.
摘要:
According to one embodiment of the present invention, a method for echocardiogram view classification is provided. According to one embodiment of the present invention, a method comprises: obtaining a plurality of video images of a subject; aligning the plurality images; using the aligned images to generate a motion magnitude image; filtering the motion magnitude image using an edge map on image intensity; detecting features on the motion magnitude image, retaining only those features which lie in the neighborhood of intensity edges; encoding the remaining features by generating, x, y image coordinates, a motion magnitude histogram in a window around the feature point, and a histogram of intensity values near the feature point; and using the encoded features to classify the video images of the subject into a predetermined classification.
摘要:
A method for inferring disease similarity by similarity retrieval of electrocardiogram time-series, comprising: acquiring user ECG waveforms correspondingly depicting many cardiac cycles of the heart of many users stored in a database; pre-processing each of the user ECG waveforms through pre-processing steps to isolate sets of single cardiac cycles corresponding to different heart-rates detected for each of the user ECG waveforms, each single cardiac cycle within the many cardiac cycles of the heart of many users corresponds to one single heart-rate detected. acquiring patient ECG waveforms depicting multiple cardiac cycles of the heart of a query patient; pre-processing the patient ECG waveforms through pre-processing steps to isolate sets of single cardiac cycles corresponding to different heart-rates detected for each of the patient ECG waveforms of the query patient, each single cardiac cycle within the multiple cardiac cycles of the heart of the query patient corresponds to one single heart-rate detected.
摘要:
A method and system of recognizing handwritten words in scanned documents, wherein by processing a document containing handwriting, features for word localization are extracted from handwritten words contained in said document through basis points taken from a single curve of text lines. The method is independent of page orientation, and does not assume that the individual lines of handwritten text are parallel, and the method does not require that word regions be aligned with text line orientation wherein intra-word statistics are derived from sample pages rather than using a fixed threshold. The method has applications in digital libraries, handwriting tokenization, document management and OCR systems.
摘要:
A method for recognizing job separator pages that tolerates more changes in imaging conditions involving pixel resolution, noise, pose changes (rotation, translation and shear) and a system which can be trained on-site to learn any new separator page. The invention is suitable for use in any scanning assembly systems such as flatbed scanners, feed through document handlers, single page feed scanners, as well wraparound document handlers, and is insensitive to leading and trailing edge-base skew detection problems. A job sheet is visually recognized. Detection of alignment markers is invariant to skew of the most general form and includes all rotations and stretches that occur on a page due to page folding.
摘要:
A method for monitoring dependent metric streams for anomalies including identifying a plurality of sets of dependent metric streams from a plurality of metric streams of a computer system by measuring an association of the plurality of metric streams using a statistical dependency measure analysis, wherein each set includes a plurality of the dependent metric streams and each metric stream includes a plurality of data, determining a subset of the plurality of sets of dependent metric streams to monitor by selecting a quantity of the sets of dependent metric streams that have a highest statistical dependency, cleaning the data of each set of dependent metric streams of the subset by identifying and removing outlier data, fitting a probability density function to the cleaned data of each set of dependent metric streams of the subset, wherein the probability density function is a likelihood function that provides a likelihood of an occurrence of the cleaned data, determining a detection threshold that is a lower threshold on the likelihood of the occurrence of the cleaned data of each set of dependent metric streams of the subset based on the likelihood function, detecting an anomaly if a likelihood of an occurrence of a new data of one of the sets of dependent metric streams of the subset is less than the detection threshold, and transmitting an alert signal in response to detecting the anomaly.
摘要:
There is provided a method for creating a geometric hash tree in a document processing system having a memory. A plurality of images are stored in the memory and organized in a database. Each image includes curve groups wherein each curve group is corresponded with a feature set. The method for creating a geometric hash tree includes the steps of: (1) associating a list of basis triples with an affine coordinate set, the basis triples and the affine coordinate set both varying as a function of the images and their corresponding curve groups; (2) storing both the affine coordinate set and the list of basis triples in the memory; (3) quantizing the affine coordinate set into a plurality of subsets; (4) assigning an order to the plurality of subsets; and (5) creating a geometric hash tree with the quantized affine coordinate set using the order from (4) such that the geometric hash tree is more compact in size than a conventional geometric hash table.