Abstract: The disclosed embodiments illustrate methods and systems for monitoring a worker performing a crowdsourced task being presented on a computing device. The method comprises performing at least one of a facial detection processing or an eye tracking processing, on a video stream captured by a camera of the computing device. An inattention instance of the worker is determined based on at least one of the facial detection processing or the eye tracking processing. Further, the inattention instance is communicated to a crowdsourcing platform, wherein the crowdsourced task is received from the crowdsourcing platform.

Abstract: A method for inspecting connectors of multi-fiber optical cables is disclosed. The method may include receiving a plurality of images of a plurality of areas of a connector and determining, via at least a processor, matching features in two or more images according to unique visual characteristics of the images. The method further may include aligning the images, according to unique matching features of the images, and combining the images such as to obtain a combined-image of the full connector or a desired area of the connector. The combined-image of the connector may be displayed on a monitor or a display.

Abstract: Embodiments of the present invention disclose methods for generating a strong classifier for face detection. The methods include determining, according to a size of a prestored image training sample, a parameter of weak classifier of the image training sample, obtaining a sketch value of each of the weak classifiers of the image training sample, calculating a weighted classification error of each of the weak classifiers according to the sketch value and an initial weight of the image training sample, obtaining at least one optimal weak classifier according to the weighted classification error, and generating a strong classifier for face detection according to the optimal weak classifiers. The embodiments of the present invention further disclose an apparatus for generating a strong classifier for face detection. The embodiments of the present invention have advantages of improving robustness of code against noise and reducing a false detection rate of face detection.

Abstract: A convolutional neural network is trained to analyze input data in various different manners. The convolutional neural network includes multiple layers, one of which is a convolution layer that performs a convolution, for each of one or more filters in the convolution layer, of the filter over the input data. The convolution includes generation of an inner product based on the filter and the input data. Both the filter of the convolution layer and the input data are binarized, allowing the inner product to be computed using particular operations that are typically faster than multiplication of floating point values. The possible results for the convolution layer can optionally be pre-computed and stored in a look-up table. Thus, during operation of the convolutional neural network, rather than performing the convolution on the input data, the pre-computed result can be obtained from the look-up table.

Abstract: A method for analyzing an image recorded by a camera of a vehicle. The method includes a step of reading the image of the camera. Furthermore, the method includes a step of recognizing at least one object in a subsection of the image, the subsection imaging a smaller area of the vehicle surroundings than the image. Furthermore, the method includes a step of transmitting the subsection and/or information about the subsection of the image to a driver assistance module. Finally, the method includes a step of using the subsection of the image and/or the information about the subsection of the image instead of the entire image in the driver assistance module, in order to analyze the image recorded by the camera.

Abstract: A method of identifying a palm area for a touch panel has steps of: receiving sensing frame information having multiple touching sensing points from the touch panel; selecting one of the touching sensing points; outwardly extending a distance from an outline of the selected touching sensing point to define a searching range; checking whether other touching sensing points are within the searching range; marking the touching sensing points in the searching range and expanding the searching range based on the currently marked touching sensing points; sequentially selecting and checking each touching sensing point if it is within the present searching range; and finally merging all the outlines of the marked touching sensing points to form a final outline as a palm area. Other unmarked touching sensing points are defined as touching reference points.

Abstract: Methods, apparatus, and other embodiments associated with classifying a region of cancerous tissue using a Histogram of Hosoya are described. One example apparatus includes a set of logics that acquires an image of a region of tissue demonstrating cancerous pathology, constructs a cell graph of the region of tissue, decomposes the cell graph into a set of subgraphs, computes a Hosoya Index for a subgraph, constructs a Histogram of Hosoya for the image based on the distribution of the subgraphs, and classifies the image based on the Histogram of Hosoya. Embodiments of example apparatus may generate and display the Histogram of Hosoya for the image. A prognosis for the patient may be provided based on the distribution of the histogram.

Abstract: A media decoding method based on cloud computing and decoder thereof are provided by embodiments of the present invention, which are easy to use and applicable to a media of any form, and its requirement for computer resource is low. The method includes: extracting representing features from a media code stream to be decoded; searching in the cloud for a media object which has similar representing features with the media code stream to be decoded by using a feature matching method and the representing features extracted; filling, replacing and improving parts or segments of the media code stream to be decoded with whole or parts of the media object.

Abstract: A protective camera enclosure (3) for a monitoring camera (1) comprises a cover portion (5) which is selectively moveable, in response to a detected tampering threat, from a retracted configuration exposing the camera (1) to an extended configuration covering and physically protecting the camera (1).

Abstract: Disclosed is an object tracking method comprising steps of obtaining a grayscale image sequence and a depth image sequence; predicting a candidate target region of a tracking object in a current image; determining a depth weight image of the candidate target region; acquiring a target template which is expressed by a probability density distribution of features; expressing, by a probability density distribution of features, a candidate target whose position moves in the candidate target region; calculating, based on a probability density expression of the target template and a probability density expression of the candidate target, a similarity between the target template and the candidate target, so as to get a confidence level image of the candidate target region; and locating, based on the confidence level image of the candidate target region, a position of the tracking object in the current image.

Abstract: The present invention relates to computation of digital fingerprint of a video sequence. The invention presents systems and methods for quick identification of shot boundaries and extraction of fingerprints by processing one or more specific frames. The systems and methods are applied on uncompressed video or compressed video having inter-frame or intra-frame compression. The methods comprises of comparing two frames of the video having a gap in between and identifying a specific frame present in between the two frames such that the specific frame may have a shot boundary. Shot boundaries are calculated for the entire video and then a fingerprint is generated using all the shot boundaries present in the video.

Abstract: A method, apparatus, system, and computer readable storage medium provide the ability to pre-segment point cloud data. Point cloud data is obtained and segmented. The segment information is stored. An indexing structure is created and instantiated with the point cloud data and the segment information. Based on the segment information, a determination is made regarding points needed for shape extraction. Needed points are fetched from the indexing structure an used to extract shapes. The extracted shapes are used to cull points from the point cloud data.

Abstract: In an infrastructure that uses a mobile order fulfillment system, robotic drive units may be dispatched and instructed to bring inventory holders to a workstation where at least one of the inventory holders may be operated upon. A random pattern or other image may be applied to a floor surface of the infrastructure, from which a map may be created that associates locations within the infrastructure to different portions of the image. A robotic drive unit may then reference the map along with images of the floor surface captured by its camera in order to determine its current location, as well as how to reach a desired destination.

Abstract: Provided is an image processing device capable of executing a determination of a display position of annotation information in consideration of a characteristic part of a target image without having a person designate a characteristic part of the target image or a part other than a characteristic part of the target image. An area setting unit sets a plurality of areas within a target image. A calculation unit calculates, as to each of the plurality of areas, a total value of edge amounts of the pixels within the area. An annotation information display area determination unit determines a display area of annotation information, based on the total value of each of the plurality of areas. A data output unit outputs data on the target image with the annotation information displayed in the display area determined by the annotation information display area determination unit.

Abstract: Methods, systems, and processor-readable media for video anomaly detection based upon a sparsity model. A video input can be received and two or more diverse descriptors of an event can be computed from the video input. The descriptors can be combined to form an event matrix. A sparse reconstruction of the event matrix can be performed with respect to an over complete dictionary of training events represented by the diverse descriptors. A step can then be performed to determine if the event is anomalous by computing an outlier rejection measure on the sparse reconstruction.

Abstract: A stereoscopic measurement system captures stereo images and determines measurement information for user-designated points within stereo images. The system comprises an image capture device for capturing stereo images of an object. A processing system communicates with the capture device to receive stereo images. The processing system displays the stereo images and allows a user to select one or more points within the stereo image. The processing system processes the designated points within the stereo images to determine measurement information for the designated points.

Abstract: Provided is a mutual-capacitance palm print identification method, a mutual-capacitance palm print identification device and a mutual-capacitance palm print identification touch panel. The method includes: forming a palm template information by using a mutual-capacitance touch panel to acquire a capacitive sensing data obtained from a palm pressing by a user; forming a current user's palm information by using the mutual-capacitance touch panel to acquire a capacitive sensing data obtained from a palm pressing by the current user; and comparing the current user's palm information with the palm template information and outputting the result of the comparison. It is unnecessary to add extra parts for the present invention, and thus the cost is low. Moreover, a palm of a living body is required, it is difficult to be cracked and counterfeited.

Abstract: A pattern processing apparatus includes an acquisition unit, an adjacent element identification unit, a connection unit, and an assigning unit. The acquisition unit acquires a pattern image in which pattern elements are arrayed in a grid, together with coordinate information of the pattern elements. The adjacent element identification unit identifies, for each of the pattern elements thus acquired, adjacent elements, in at least two directions, each having the shortest distance to the each of the pattern elements. The distances between the each of the pattern elements and other pattern elements are calculated using the coordinate information. The connection unit connects and groups, as a connected component, the pattern elements that are adjacent to one another. The assigning unit assigns position information in the array in order from the pattern element which is a starting point, among the pattern elements included in the connected component, according to a predetermined regularity.

Abstract: A stereoscopic measurement system captures stereo images and determines measurement information for user-designated points within stereo images. The system comprises an image capture device for capturing stereo images of an object. A processing system communicates with the capture device to receive stereo images. The processing system displays the stereo images and allows a user to select one or more points within the stereo image. The processing system processes the designated points within the stereo images to determine measurement information for the designated points.

Abstract: A video image capture component includes a light source operable in a first spectrum, a first image detector operable in the first spectrum, a second light source operable in a second spectrum, and a second image detector operable in the second spectrum. A filtering component generates a combination image by filtering a first image obtained by the first image detector with a high-contrast filter, resulting in a high-contrast image, and masking a second image obtained by the second image detector using the high-contrast image. A compositing component creates a composite image from the combination image and a selected image. A display component displays the composite image. Alternative systems and methods for creating a combination image include techniques involving thermal imaging, laser detection, and narrow band frequency detection.