2391 results about "Visual recognition" patented technology

The invention relates to a visual recognition and positioning method for robot intelligent capture application. According to the method, an RGB-D scene image is collected, a supervised and trained deep convolutional neural network is utilized to recognize the category of a target contained in a color image and a corresponding position region, the pose state of the target is analyzed in combinationwith a deep image, pose information needed by a controller is obtained through coordinate transformation, and visual recognition and positioning are completed. Through the method, the double functions of recognition and positioning can be achieved just through a single visual sensor, the existing target detection process is simplified, and application cost is saved. Meanwhile, a deep convolutional neural network is adopted to obtain image features through learning, the method has high robustness on multiple kinds of environment interference such as target random placement, image viewing anglechanging and illumination background interference, and recognition and positioning accuracy under complicated working conditions is improved. Besides, through the positioning method, exact pose information can be further obtained on the basis of determining object spatial position distribution, and strategy planning of intelligent capture is promoted.

A humanoid robot equipped with an interface for natural dialog with an interlocutor is provided. Previously, the modalities of dialog between humanoid robots equipped moreover with evolved displacement functionalities and human beings are limited notably by the capabilities for voice and visual recognition processing that can be embedded onboard said robots. The present disclosure provides robots are presently equipped with capabilities to resolve doubt on a several modalities of communication of the messages that they receive and for combining these various modalities which make it possible to greatly improve the quality and the natural character of dialogs with the robots' interlocutors. This affords simple and user-friendly means for carrying out the programming of the functions making it possible to ensure the fluidity of these multimodal dialogs.

A liquid crystal display device has first and second liquid crystal panels (101, 102) mainly comprising liquid crystal cells (107, 108), respectively, disposed back to back to enable the visual recognition of the liquid crystal panels (101, 102). A light guide plate (112) is disposed between the first liquid crystal panel (101) and the second liquid crystal panel (102) with a light source (114) located adjacent to its one end face (112c), and a polarization separator (110) located between the first liquid crystal panel (101) and the light guide plate (112). Light emitted from the light guide plate (112) is split into two beams of polarized light: one is emitted to the first liquid crystal panel (101), and the other to the second liquid crystal panel (102) through the light guide plate (112). This constitution thins down a both-sided display liquid crystal display device and reduces power consumption.

An apparatus and method for analyzing known data, storing the known data in a pattern database (“PDB”) as a template is provided. Additional methods are provided for comparing new data against the templates in the PDB. The data is stored in such a way as to facilitate the visual recognition of desired patterns or indicia indicating the presence of a desired or undesired feature within the new data. The apparatus and method is applicable to a variety of applications where large amounts of information are generated, and/or if the data exhibits fractal or chaotic attributes.

The invention discloses a material object programming method and a material object programming system, which belong to the field of human-machine interaction. The method comprises the following steps of: 1) establishing a set of material object programming display environment; 2) shooting the sequence of material object programming blocks which are placed by a user and uploading the shot image toa material object programming processing module by using an image acquisition unit; 3) converting the sequence of the material object blocks into a corresponding functional semantic sequence by usingthe material object programming processing module according to the computer vision identification modes and the position information of the material object programming blocks; 4) determining whether the current functional semantic sequence meets the grammatical and semantic rules of the material object display environment or not, and if the current functional semantic sequence does not meet the grammatical and semantic rules of the material object display environment, feeding back a corresponding error prompt; 5) replacing the corresponding material object programming blocks by using the useraccording to the prompt information; and 6) repeating the steps 2) to 5) until the functional semantic sequence corresponding to the sequence of the placed material object programming blocks meets the grammatical and semantic rules of the material object display environment, and finishing a programming task. By using the method and the system, the problem that children and green hands are difficult to learn programming is solved, and the system has low cost and is easy to popularize.

Around-vehicle visual recognition apparatus, methods, and programs acquire circumstances of a vehicle, including the vehicle's location and cause a camera to take images around the vehicle. The apparatus, methods, and programs determine at least one first area around the vehicle that is more relevant to a driver based on the circumstances of the vehicle than another area around the vehicle and display an image of the at least one first area on a display.

The invention relates to a fine granularity vehicle multi-property recognition method based on a convolutional neural network and belongs to the technical field of computer visual recognition. The method comprises the steps that a neural network structure is designed, including a convolution layer, a pooling layer and a full-connection layer, wherein the convolution layer and the pooling layer areresponsible for feature extraction, and a classification result is output by calculating an objective loss function on the last full-connection layer; a fine granularity vehicle dataset and a tag dataset are utilized to train the neural network, the training mode is supervised learning, and a stochastic gradient descent algorithm is utilized to adjust a weight matrix and offset; and a trained neural network model is used for performing vehicle property recognition. The method can be applied to multi-property recognition of a vehicle, the fine granularity vehicle dataset and the multi-propertytag dataset are utilized to obtain more abstract high-level expression of the vehicle through the convolutional neural network, invisible characteristics reflecting the nature of the to-be-recognizedvehicle are learnt from a large quantity of training samples, therefore, extensibility is higher, and recognition precision is higher.

A system detects a transaction outcome by obtaining video data associated with a transaction area and by obtaining transaction data concerning at least one transaction that occurs at the transaction area. The system correlates the video data associated with the transaction area to the transaction data to identify specific video data captured during occurrence of that at least one transaction at the transaction area. Based a transaction classification indicated by the transaction data, the system processes the video data to identify appropriate visual indicators within the video data that correspond to the transaction classification.

A visual recognition and identification apparatus comprising a mounting means (2, 18) adapted for placement in a physical environment wherein the mounting means incorporates one or a plurality of light emitting diodes (1) adapted to provide a visual signal characterised in the light emitting diodes (1) are mounted and physically associated with the printed circuit circuit board (2, 18) wherein the printed circuit board (2, 18) has been modified to provide a distinct angle mounting for the light emitting diodes (1) to provide a defined viewing angle and wherein the printed circuit board (2, 18) is manufactured from the compliant material so as to allow the visual recognition and identification apparatus to be moulded and adapted to a number of environments including fixture to non flat surfaces, incorporation into clothing (19, 21), caps and badges by way of sewing or incorporation into a range of devices.

The invention discloses an accurate visual positioning and orienting method for a rotor wing unmanned aerial vehicle on the basis of an artificial marker. The accurate visual positioning and orienting method includes the following steps that the marker with a special pattern is installed on the surface of an artificial facility or the surface of a natural object; a camera is calibrated; the proportion mapping relation among the actual size of the marker, the relative distance between the marker and the camera and the size, in camera imaging, of the marker is set up, and the keeping distance between the unmanned aerial vehicle and the marker is set; the unmanned aerial vehicle is guided to fly to the position where the unmanned aerial vehicle is to be suspended, the unmanned aerial vehicle is adjusted so that the marker pattern can enter the visual field of the camera, and a visual recognition function is started; a visual processing computer compares the geometrical characteristic of the pattern shot currently and a standard pattern through visual analysis to obtain difference and transmits the difference to a flight control computer to generate a control law so that the unmanned aerial vehicle can be adjusted to eliminate the deviation of the position, the height and the course, and accurate positioning and orienting suspension is achieved. The accurate visual positioning and orienting method is high in independence, good in stability, high in reliability and beneficial for safety operation, nearby the artificial facility the natural object, of the unmanned aerial vehicle.

Systems and methods for presenting search results based on relevancy in an interactive media guidance application are disclosed. After performing a user-initiated or automatic search for media content, the interactive media guidance application determines which of the hits are most relevant to the user. The guidance application then displays, or visually identifies, the relevant items. Some embodiments employ using different display arrangements based on the number of relevant items. Some embodiments display the relevant items in recommendation lists or hot lists.

An apparatus and method for analyzing known data, storing the known data in a pattern database (“PDB”) as a template is provided. Additional methods are provided for comparing new data against the templates in the PDB. The data is stored in such a way as to facilitate the visual recognition of desired patterns or indicia indicating the presence of a desired or undesired feature within the new data. Data may be analyzed as fragments, and the characteristics of various fragments, such as sting length, may be calculated and compared to other indicia to indicate the presence or absence of a particular substance, such as a hydrocarbon. The apparatus and method is applicable to a variety of applications where large amounts of information are generated, and/or if the data exhibits fractal or chaotic attributes.

The invention relates to an on-line detection grading device for a lithium ion battery pole piece and a method thereof. In the device, image data of the lithium ion battery pole piece are acquired by using an embedded visual recognition device and are analyzed and processed to judge whether the appearance quality of the pole piece meets requirements or not, so that the actions of a turnover part and a grading part are controlled to realize real-time on-line double-side detection and grading of the electrode pole piece. The method has high detection speed; generally, the detection of the frontside and the back side of one electrode pole piece can be finished within 1 second, so that the production efficiency is increased; if the pole piece is manually detected, longer time is required to be taken; the detection accuracy is high; the embedded visual detection technology is adopted, so that the accuracy is high and the stability is also realized, and a detection result has high uniformity; the running time is long, namely a detection system can uninterruptedly run for 24 hours, so that the productivity is improved and the labor cost is saved; and on-line detection can be realized, so that the on-line detection of the electrode pole piece is realized, the interference in a manual operation process is reduced, and the productivity is improved.

An apparatus and method for analyzing known data, storing the known data in a pattern database (“PDB”) as a template is provided. Additional methods are provided for comparing new data against the templates in the PDB. The data is stored in such a way as to facilitate the visual recognition of desired patterns or indicia indicating the presence of a desired or undesired feature within the new data. Data may be analyzed as fragments, and the characteristics of various fragments, such as string length, may be calculated and compared to other indicia to indicate the presence or absence of a particular substance, such as a hydrocarbon. The length and/or character of each fragment are a product of the cutting criteria in both horizontal and vertical orientations. Modifying the cutting criteria can have a beneficial effect on the results of the analysis and/or in the amount of time necessary to achieve useful results. The apparatus and method is applicable to a variety of applications where large amounts of information are generated, and/or if the data exhibits fractal or chaotic attributes.

The invention discloses a target detection and recognition method for water surface panoramic images, and belongs to the technical field of computer visual recognition. The method comprises the following steps of: firstly carrying out target detection and recognition on a panoramic image by utilizing a target detection and recognition neural network model, and obtaining a target type and an initial position of a detection box; carrying image segmentation on a target local area to obtain a plurality of superpixel blocks, and combining the superpixel blocks by adoption of a multi-threshold fusion strategy so as to obtain an adjusted target detection box; calculating an edge map of a new target local area, adjusting the target detection box according to peripheral edge points of the target detection box so as to obtain a final target detection box; and finally converting the position of the final target detection box into a target practical position. According to the method, distortion effect in panoramic images can be effectively overcome, and target positions can be correctly recognized from the panoramic images.

Visually identifying one or more known or anonymous voice speakers to a listener in a computing session. For each voice speaker, voice data include a speaker identifier that is associated with a visual indicator displayed to indicate the voice speaker who is currently speaking. The speaker identifier is first used to determine voice privileges before the visual indicator is displayed. The visual indicator is preferably associated with a visual element controlled by the voice speaker, such as an animated game character. Visually identifying a voice speaker enables the listener and/or a moderator of the computing session to control voice communications, such as muting an abusive voice speaker. The visual indicator can take various forms, such as an icon displayed adjacent to the voice speaker's animated character, or a different icon displayed in a predetermined location if the voice speaker's animated character is not currently visible to the listener.

Viewability is improved creating more reliable visual recognition of an image including a copy-forgery-inhibited pattern image to clearly distinguish a copy and an original being displayed for preview. An image processing apparatus produces data of a copy-forgery-inhibited pattern image that is added to an image to be output for printing and comprises a latent image and a background image. The apparatus includes: a display unit for displaying an image; and a display control unit for distinctively displaying images, on the display unit, in a first display state displaying the copy-forgery-inhibited pattern image and in a second display state in which a display mode of at least one of the latent image and the background image of the copy-forgery-inhibited pattern image differs from a display mode of the image displayed in the first display state.

An object is to provide an artificial vision system ensuring a wide field of view without damaging a retina. In the artificial vision system, a plurality of electrodes (23) are to be implanted so as to stick in an optic papilla of an eye of a patient. A signal for stimulation pulse is generated based on an image captured by an image pick up device (11) to be disposed outside a body of the patient. The electrical stimulation signals outputted from the electrodes (23) based on the signals for stimulation pulse stimulate an optic nerve of the eye, thereby enabling the patient to visually recognize the image from the image pickup device (11).

A technique is provided for improving visual recognition of letters, graphics, etc., in a portion of a displayed image that has been enlarged. An image display system includes an image display component for displaying on a screen an image based on image data, a region indicator component for permitting a region to be enlarged from the image to be selected and identified, a data modification component for modifying the portion of the image data corresponding to an enlarged display region so that the selected image is enlarged and displayed within the enlarged display region on the screen, and a brightness adjusting component for adjusting values instructing brightness of the corresponding image data part such that a contrast of the image within the enlarged display is higher than a contrast of the original image within the enlarged display region.

Disclosed is an operation input system including: an operation display device; and a display device, wherein in case that the operation display device receives input of security information, the operation display device sets input keys in the key arrangement on a screen of the display unit and displays the operation window in which the input keys or at least contents of the input keys cannot be visually recognized, and the operation detecting unit receives the operation for the set input keys, and the display device controls display contents of the second display unit so as to display a state in which a virtual image of the input keys is combined with a situation in which an operator operates the display unit in accordance with an image shot by the camera so as to match a position of the virtual image with the set input keys.