306 results about "Vision algorithms" patented technology

A method to identify character strings associated with potentially malicious software items. The method includes employing a visual algorithm to translate one or more characters of a character string into corresponding characters in a visual ID for use in grouping and comparing computer items having similar visual IDs, such as a reference ID for a computer item that is known to be non-malicious. The method may, among other things, elucidate an attacker's attempt to obfuscate malicious software by using file names that are very similar to those used for harmless files.

The invention discloses a man-machine interactive system and a real-time gesture tracking processing method for the same. The method comprises the following steps of: obtaining image information of user side, executing gesture detection by a gesture detection unit to finish separation of a gesture and a background, and automatically determining a smaller rectangular frame which surrounds the hand as a region of interest in the image information through a vision algorithm; calculating the hand profile state of each frame in a video sequence by a gesture tracking unit; executing validity check on hand actions according to the calculated hand profile state, and determining the gesture action finished by the user; and generating a corresponding gesture action control instruction according to the determined gesture action, and making corresponding feedbacks by a three-dimensional user interface according to the gesture action control instruction. In the system and the method, all or part of gesture actions of the user are sensed so as to finish accurate tracking on the gesture, thus, a real-time robust solution is provided for an effective gesture man-machine interface port based on a common vision sensor.

The invention relates to an unmanned aerial vehicle aerial photography sequence image-based slope three-dimension reconstruction method. The method includes the following steps that: feature region matching and feature point pair extraction are performed on un-calibrated unmanned aerial vehicle multi-view aerial photography sequence images through adopting a feature matching-based algorithm; the geometric structure of a slope and the motion parameters of a camera are calculated through adopting bundle adjustment structure from motion and based on disorder matching feature points, and therefore, a sparse slope three-dimensional point cloud model can be obtained; the sparse slope three-dimensional point cloud model is processed through adopting a patch-based multi-view stereo vision algorithm, so that the sparse slope three-dimensional point cloud model can be diffused to a dense slope three-dimensional point cloud model; and the surface mesh of the slope is reconstructed through adopting Poisson reconstruction algorithm, and the texture information of the surface of the slop is mapped onto a mesh model, and therefore, a vivid three-dimensional slope model with high resolution can be constructed. The unmanned aerial vehicle aerial photography sequence image-based slope three-dimension reconstruction method of the invention has the advantages of low cost, flexibility, portability, high imaging resolution, short operating period, suitability for survey of high-risk areas and the like. With the method adopted, the application of low-altitude photogrammetry and computer vision technology to the geological engineering disaster prevention and reduction field can be greatly prompted.

The present invention provides a method for reliably detecting change in the 3-D shape of objects. It uses an estimate of the accuracy of the 3-D models derived from a set of images taken simultaneously. This accuracy estimate is used to distinguish between significant and insignificant changes in 3-D models derived from different image sets. In one embodiment of the present invention, the accuracy of the 3-D model is estimated using self-consistency methodology for estimating the accuracy of computer vision algorithms. In another embodiment of the present invention, resampling theory is used to compare the mean or median elevation for each change in the models. This methodology allows for estimating, for a given 3-D reconstruction algorithm and class of scenes, the expected variation in the 3-D reconstruction of objects as a function of viewing geometry and local image-matching quality (referred to as a “score”). Differences between two 3-D reconstructions of an object that exceed this expected variation for a given significance level are deemed to be due to a change in the object's shape, while those below this are deemed to be due to uncertainty in the reconstructions.

A system for monitoring shopping baskets (e.g., baskets on human-propelled carts, motorized carts, or hand-carried baskets) can include a computer vision unit that can image a surveillance region (e.g., an exit to a store), determine whether a basket is empty or loaded with merchandise, and assess a potential for theft of the merchandise. The computer vision unit can include a camera and an image processor programmed to execute a computer vision algorithm to identify shopping baskets and determine a load status of the basket. The computer vision algorithm can comprise a neural network. The system can identify an at least partially loaded shopping basket that is exiting the store, without indicia of having paid for the merchandise, and execute an anti-theft action, e.g., actuating an alarm, notifying store personnel, activating a store surveillance system, activating an anti-theft device associated with the basket (e.g., a locking shopping cart wheel), etc.

The invention relates to a multipoint touch method based on binocular stereoscopic vision, which relates to the technical field of computer vision. The method comprises the following steps: firstly, the azimuth and the projection parameters of a camera relative to a three-dimensional scene are determined through camera calibration, and a binocular camera is utilized to acquire an image pair; secondly, two pictures on the same frame which are acquired on the first step are taken as input, and a scene depth map is calculated by the stereoscopic vision algorithm; thirdly, closest patches are obtained through Gaussian blur and determined thresholds, and central coordinates of the patches are calculated; and fourthly, a plurality of coordinates obtained in the third step are taken as input of a multipoint touch program, and interactive results are returned to a screen. The multipoint touch method based on binocular stereoscopic vision can realize completing the multipoint touch operation in space by a user without touching the screen, and coordinate information of input points acquired is three-dimensional, so that the multipoint touch method can realize more complex interaction such as push and pull of a virtual object and so on.

The invention relates to a method for measuring coal bunker material level based on laser and binocular vision and a device thereof, belonging to the field of machine vision. The device comprises two CCD cameras with the same parameters, two lenses with consistent parameters, one or a plurality of lasers, a light source, a processing element and an explosion-proof case provided with a window. A rotating mechanism is also included when one laser is used. The method includes that the laser emits variable size laser beams, so as to form a light spot on coal surface, image parallax of the light spot between the two cameras is utilized to measure material level, and an image pair is acquired by the two cameras; corrosion expansion pre-treatment is carried out on the obtained images by utilizing an improved algorithm; for the two images after being pretreated, a new rapid stereoscopic vision algorithm is utilized to calculate a depth map; depth is obtained according to the depth map, and finally high efficiency, real time and accurate coal bunker material level measurement is realized. The device has two measuring modes, namely real-time detection mode and sensor detection mode.

The invention provides a visual SLAM algorithm based on instance segmentation, and the algorithm comprises the steps: firstly extracting feature points of an input image, and carrying out the instancesegmentation of the image through employing a convolutional neural network; secondly, using instance segmentation information for assisting in positioning, removing feature points prone to causing mismatching, and reducing a feature matching area; and finally, constructing a semantic map by using the semantic information segmented by the instance, thereby realizing reuse and man-machine interaction of the established map by the robot. According to the method, experimental verification is carried out on image instance segmentation, visual positioning and semantic map construction by using a TUM data set. Experimental results show that the robustness of image feature matching can be improved by combining image instance segmentation with visual SLAM, the feature matching speed is increased,and the positioning accuracy of the mobile robot is improved; and the algorithm can generate an accurate semantic map, so that the requirement of the robot for executing advanced tasks is met.

In trains a dragging air hose can snag an object on the tracks and break. A broken air hose can lead to an emergency stop or an accident caused by faulty brakes. The present invention mounts a video camera opposite a lit screen, thereby contrasting the air hose profile against a white screen in all weather and lighting conditions. Machine vision algorithms locate the air hose in a captured image from a moving train, and compute its lowest point. An alarm condition is sent to a remote location. An optional car gap detector is laser based. It helps align the car gap with the data capture for the camera. Another subsystem can include a car coupler height detection algorithm.

The subject disclosure is directed towards a hybrid stereo image / motion parallax system that uses stereo 3D vision technology for presenting different images to each eye of a viewer, in combination with motion parallax technology to adjust each image for the positions of a viewer's eyes. In this way, the viewer receives both stereo cues and parallax cues as the viewer moves while viewing a 3D scene, which tends to result in greater visual comfort / less fatigue to the viewer. Also described is the use of goggles for tracking viewer position, including training a computer vision algorithm to recognize goggles instead of only heads / eyes.

The embodiment of the invention provides an aircraft autonomous obstacle avoidance system and method and an aircraft. A binocular camera and a millimeter wave radar are combined. A push broom stereo vision algorithm and an advanced image data fusion method are used to mark an obstacle and detect the obstacle. The position and shape and texture information of the obstacle in front of the aircraft can be quickly and accurately acquired. An accurate basis is provided for the aircraft to avoid the obstacle in real time. The autonomous flight and intelligence level of the aircraft is improved. Theenvironmental adaptability of the aircraft is enhanced. The aircraft can safely fly to complete a task in a complex environment.

The invention relates to an intelligent display system automatically tracking a head posture. The system can capture the head posture of a human body in real time, a display follows the head of the human body, the display and the head are constantly kept at optimum positions, eye fatigue can be relieved, and myopia can be prevented. The system comprises an image acquisition module, a vision algorithm processing module and a display control module. The image acquisition module is used for acquiring the image of the head of the human body in real time; the vision algorithm processing module is used for firstly preprocessing the image, constantly keeping the head at a vertical position in the image, secondly extracting feature points of a human face by the aid of an ASM (active shape model) algorithm, and finally acquiring the spatial posture of the head according to the triangular surveying principle of elevation angles and side-tipping angles. Head posture information is transmitted to a steering gear control module and then calculated by a single chip microcomputer to form a PWM (pulse width modulation) signal for controlling a steering gear, so that the display follows the head posture at two degrees of freedom of elevation and side-tipping.

Methods and systems are disclosed for real-time tracking of a multiplayer ball game using a mobile computing device. The methods and systems are configured to receive an input video of the multiplayer ball game captured using a camera on the mobile computing device in a gaming area associated with n players, where n is an integer and n≥2; detect a plurality of player postures, by performing a computer vision algorithm on each of a plurality of frames of the input video; extract an associated player feature from each of the plurality of player postures; assign each of the plurality of player postures, based on the associated player feature, to one of at least n posture groups, where each player is represented by one of the at least n posture groups; and determine a player location for each player, based on the player's posture group.

The invention relates to an extraction method and an extraction device for a robot spraying track based on curve three-dimensional reconstruction. The purpose of the invention is that the provided method and the provided device can accurately extract the spraying track and have the characteristics of high work efficiency and high adaptability. The technical scheme is that the extraction method for the robot spraying track based on curve three-dimensional reconstruction comprises the steps of S1, calibrating an intrinsic parameter matrix K and a distortion coefficient of a camera; S2, acquiring an image of a workpiece; S3, processing the image; S4, finding a group of requirement conforming homonymic curves to act target curves; and S5, carrying out three-dimensional reconstruction on the target curves by using a stereoscopic vision algorithm. The extraction device for the robot spraying track based on curve three-dimensional reconstruction comprises a conveying belt, proximity switches and a computer; and the computer performs electric signal interaction with a motion control card which is connected with a stepping motor and the proximity switches, a camera connected with an image acquisition card and an industrial robot.

The invention provides a real-time binocular vision guidance method for an underwater vehicle. Aimed at the characteristics of the marine environment where the underwater vehicle operates, the invention puts forward a pyramid normalized covariance binocular vision algorithm based on the texture control to realize the real-time property of the application level with the precision stabled at the centimeter level. Regarding the final demonstration of an environmental message, the real-time binocular vision guidance method puts forward a virtual sonar model, thereby expressing the 2.5-dimensinal message as a barrier (or target) matrix based on the virtual sonar model. And the matrix contains the depth and bearing information of the barrier (or target).

A method and system that performs low level fusion of Radar or LiDAR data with an image from a camera. The system includes a radar-sensor, a camera, and a controller. The radar-sensor is used to detect a radar-signal reflected by an object in a radar-field-of-view. The camera is used to capture an image of the object in a camera-field-of-view that overlaps the radar-field of view. The controller is in communication with the radar-sensor and the camera. The controller is configured to determine a location of a radar-detection in the image indicated by the radar-signal, determine a parametric-curve of the image based on the radar detections, define a region-of-interest of the image based on the parametric-curve derived from the radar-detection, and process the region-of-interest of the image to determine an identity of the object. The region-of-interest may be a subset of the camera-field-of-view.

The invention relates to an ROS-based autonomous mobile grabbing system and method. The system can be divided into an image acquisition module, a visual algorithm processing module, a pose detection module and a grabbing control module. According to the method, the functions of environment perception, visual navigation and real-time object pose estimation and grabbing of a mobile robot in a strange environment are achieved. The mobile robot is combined with machine vision, so that the robot can complete more intelligent tasks, good services are provided, and unnecessary burdens are relieved. The system and method have the advantages of high efficiency, high detection precision, real-time detection, good adaptability and the like, can release human input in daily work to a great extent, andhave strong market potential. Work and life of people are greatly facilitated, and improvement of the living standard is facilitated.

The invention discloses a conveying belt deviation fault detection method based on an image identification technology. The method comprises the following steps of a preparation process: collecting a monitoring image, and calibrating belt edges on a left side and a right side under a normal running situation of the conveying belt, establishing a virtual graduated scale on a calibrated position, andonly executing the processing one time in an equipment installation stage; and a detection process: in an equipment running stage, obtaining the practical positions of edges on the left side and theright side of the conveying belt through an image vision algorithm, comparing the practical position of the edge of the belt on the image and the calibration position in a preparation stage to obtainthe deviation pixel distance of the belt on a video image, and finally, reducing the deviation pixel distance into a true pixel distance through a fixed proportional scale. By use of the conveying belt deviation fault detection method, whether the belt deviates or not can be more accurately detected, and an accurate value of a deviation value can be given, the robustness of an algorithm can be improved while belt deviation accuracy is improved, so that the conveying belt deviation fault detection method can be favorably suitable for complex working conditions under an indoor condition, an outdoor condition and different illumination conditions, learning samples are not required, and sample collection cost is lowered.

The invention discloses a bed and chair split type nursing bed with butting mechanisms and an automatic butting method. The bed and chair split type nursing bed comprises a wheelchair portion and a bed portion. The wheelchair portion can be switched over between a flatly spread state and a wheelchair state, and butting guide wheels are arranged on two sides of the wheelchair portion; an open groove for accommodating the wheelchair portion is formed in the middle of the bed portion, and butting guide rails are arranged on two sides of the open groove; the wide distance between the two butting guide rails on the two sides of the open groove becomes narrow and then keeps unchanged in the direction in which the wheelchair portion moves into the open groove. The bed and chair split type nursingbed and the automatic butting method have the advantages that the open groove for accommodating a wheelchair is formed in the middle of the bed portion, the bed and chair split type nursing bed comprises butting portions with the butting guide wheels and the butting guide rails, and accordingly butting operation becomes simple; the bed and chair split type nursing bed is reasonable in structuraldesign, patients are free of falling risks during separating and butting, the wheelchair portion and the bed portion can be automatically butted to each other by vision algorithms by the aid of the automatic butting method, accordingly, the bed and chair split type nursing bed does not need to be operated by medical staffs, medical resources can be saved, and the autonomous action capacity of thepatients can be improved.

The invention provides an image analysis method based on a machine learning algorithm. Main contents comprise attribute learning, scale-based attribute metering and consistent visualization in a cluster mode. Attribute characteristics of an object of an image are provided according to depth learning of a data set carried out by a classifier. A monitoring learning method is utilized, picture notes are learned by a training machine, an object attribute classifier is established, test performance standards are improved to a maximum degree, machine classification is carried out, operation to a bigger database is further carried out, and the classifier is made to be an effective tool for picture analysis. The method is advantaged in that two novel vision algorithms are utilized, relevant attribute clusters are detected through image clustering, according to the identified attributed clusters, pictures and object clusters are further identified, the object attributes are provided according to machine learning, and the method further contributes to new design in the industrial field and innovation solution schemes in the engineering field.

The invention discloses an intelligent guiding type mechanical arm vision system which comprises a mechanical arm, an industrial camera and an industrial personal computer. Due to the fact that a camera photographing function module, an image processing function module, an industrial personal computer and mechanical arm communication function module and an industrial camera and mechanical arm fast configuration function module are installed in the industrial personal computer based on NI Vision Builder platform development, the functions of controlling the industrial camera to conduct photographing, processing obtained workpiece images, conducting communication between the industrial personal computer and the mechanical arm and fast configuring of the industrial camera and the mechanical arm are achieved. A universal communication module based on the Ethernet is adopted, so that combination of various mechanical arm communication protocols of different plants and a vision system is broken through; the industrial camera and mechanical arm fast configuration function module is adopted for establishing the fast and convenient guiding type mechanical vision system; and the image processing function module can be adopted for conducting recognition and establishing vision algorithms on a single or multiple different objects, the complex process is simplified, the processing speed is high, and the intelligence degree is high.

A stature measuring method includes the first step of training a human face classifier, the second step of training a human face and stature model, the third step of detecting a human face through the classifier, the fourth step of obtaining image data of the target human face and the fifth step of inputting the image data into the model to obtain the stature of a human body. The stature measuring method is effectively integrated through a machine learning algorithm and a vision algorithm, can be used for measuring stature in a non-contact mode under the circumstance that the distance between a user and a camera is unfixed, and guarantees security of stature data measured on some special occasions. The method breaks through restrictions on a traditional measuring tool, is convenient to operate, capable of effectively saving cost, manpower and material resources and high in transportability, and can be integrated with a variety of intelligent products to create certain economic benefits and practical value.

The present application is directed to a system and method providing vision algorithms for identifying objects traveling in space and identifying the configuration of one or more targets of the objects. The system is suitably programmed to record data related to movement of objects in relation to one or more targets.

The invention provides a depth camera based intelligent visitor flowrate detecting method for study rooms of a library, and a machine vision algorithm is combined in the method. The method can be used to count used and vacant seats in each study room in real time accurately, statistical data is displayed on a large display screen or a WeChat official account of the library, so that the number of used seats and the number of vacant seats in each study room are clear. Students who would like to study in the library can know the conditions in real time conveniently, and time waste is avoided. The device and method can improve the service quality of the library, and improves the effective utilization rate of limited educational resources.

The invention discloses AI (Artificial Intelligence) navigation glasses for blind people. The glasses consist of a glasses frame and a PCBA arranged on the glasses frame. The PCBA is provided with a binocular camera, a six-axis IMU sensor, an embedded AI processor, a voice processor, a Bluetooth / WiFi module and a mobile data transmission module. The binocular camera and the 6-axis IMU sensor collect video images and motion tracks of surroundings, and the AI navigation processor completes environment characteristic recognition, space matching, deep information extraction and three-dimensional reconstruction through a vision algorithm, and helps blind people establish environment three-dimensional perception through voice description of the environment characteristics. The voice description is output to the blind people through the bone conduction earphone, and the blind people wearing the glasses complete navigation and avoid obstacles under the prompt of the voice so as to realize free walking.

The invention discloses a robot orientation and hand-eye relationship simultaneous calibration method based on bundle adjustment, which comprises the steps of performing three-dimensional reconstruction on a detection target through a monocular stereoscopic vision algorithm to obtain a camera external parameter matrix without a scale factor alpha; utilizing a robot kinematics equation to obtain arobot motion orientation matrix; representing a calibration equation A(alpha)X=ZB by using a matrix direct product parameter method, and solving a robot orientation and hand-eye relationship initial value; and carrying out iterative optimization on the robot orientation and hand-eye relationship initial value and a three-dimensional reconstruction result obtained by the monocular stereoscopic vision algorithm by using a beam adjustment optimization model. According to the robot orientation and hand-eye relationship simultaneous calibration method based on beam adjustment, based on the light beam adjustment model, accurate three-dimensional positioning can be performed on the target while the robot orientation and the hand-eye relationship are solved, accurate positioning of the detection target is also realized while dependence on the target is eliminated, and thus the intelligent robot is endowed with sensing and positioning functions.

An image of a region captured by an unmanned aerial vehicle flying at an altitude may be received. A computer vision algorithm may be executed with the image as an input to compute an overall confidence score associated with detecting one or more candidate objects in the image. Responsive to determining that the confidence score is below a predefined minimum threshold or above a predefined maximum threshold, the unmanned aerial vehicle may be controlled to change its altitude and recapture the image of the region at a new position. Responsive to determining that the overall confidence score is not below the predefined minimum threshold, information associated with the image may be stored on a storage device.

The invention provides a handheld interactive device and a projection interactive method thereof. The handheld interactive device comprises a projection module, a photographing module, a sensing control module, a central processing unit, a wireless communication module and a storage module. The projection module is used for projecting initial virtual projection information into a real projection space. The photographing module is used for collecting virtual and real projection space image data information, establishing a coordinate transformation model and obtaining coordinate transformation parameter information. The sensing control module is used for obtaining the relative position information of the handheld interactive device and an initial real projection space. The central processing unit is used for receiving data information from the photographing module, the projection module and the sensing control module, processing and analyzing the data information according to an image vision algorithm and controlling the projection module to project corresponding virtual projection information according to analyzing results. The projection module, the photographing module, the sensing control module, the wireless communication module and the storage module are electrically connected with the central processing unit and are controlled by the central processing unit.

A universal UAV (unmanned aerial vehicle) vision simulation platform comprises a simulation environment part and a real environment part, wherein the simulation environment part comprises an ROS (robot operating system), an open-source robot simulation platform module Gazebo, a vision algorithm library and a universal UAV connecting port for converting information transmitted by the ROS into control information capable of being recognized by different flight controllers, wherein the universal UAV connecting port supports the ROS; the open-source robot simulation platform module Gazebo, the vision algorithm library and the universal UAV connecting port are connected with the ROS; the real environment part comprises an aircraft rack and a flight controller, wherein the universal UAV connecting port is connected with the flight controller. The UAV vision simulation platform is good in expansibility and universality.

A double-camera machine vision positioning method of a surface mounted component includes a process of using two cameras to collect images of the surface mounted component, a tray platform which is translatable and rotatable in a horizontal direction, two cameras are arranged above the tray platform, an image data transmission channel is arranged between the cameras and a computer, a calibration step is carried out at first, and then a step of positioning a workpiece to be machined is carried out. The two cameras are used to shoot images, and a series of positioning motion data are collected, thereby obtaining a coordinate system of positioning by two points collected by vision, two points on the workpiece which have a certain distance are sampled, the measuring cardinal number is enlarged by one order of magnitude, errors of the images can almost be ignored, dependency on shapes of feature points and a vision algorithm can be avoided, and thus the double-camera machine vision positioning method of the surface mounted component has a wider range of application, and is higher in precision.