196 results about "Face size" patented technology

The invention relates to a method and system for screening face images. The method comprises the following steps: according to preset face integrity standards, face size threshold and face position threshold, obtaining a first screening result; according to a preset definition threshold, obtaining a second screening result; according to a preset best angle threshold, obtaining a third screening result; and according to the sum of similarities between each face image and other face images, the definition of each face image and the face deflection angle, obtaining the weighting sum of each face image in the third screening result, and acquiring a face image with the maximum weighting sum for face identification. The method can be used for solving the problem that the face is mistakenly identified due to the low quality of an input image, the data volume of the face images is effectively reduced, and the operation load during face identification is relieved; and moreover, the data volume of the face images is reduced, front face images with high quality are effectively screened to be identified, and the accuracy and real-time performance of face identification are improved.

A controller controls a photographing angle of view based on the face size detected by the face size calculation unit and the face inclination detected by the face inclination detection unit, wherein the controller controls the photographing angle of view so that a size with respect to the photographing angle of view of the subject image recognized by the face recognition unit becomes a predetermined size.

An imaging apparatus includes a face detecting unit detecting a face region from an input image input to the imaging apparatus and a control unit calculating a distance to a subject on the basis of the face size detected by the face detection unit and displaying distance identification information indicating a distance corresponding to the calculated distance in a display unit.

A hydraulic pressure compensation system for valve actuator assemblies is described having particular application for subsea wellhead installations. The compensation system includes at least one valve actuator assembly having a housing that retains a reciprocable piston therewithin. The piston is spring biased into its fail safe configuration. The valve actuator assembly is hydraulically associated with an accumulator reservoir that defines a closed fluid reservoir and an open fluid reservoir that is exposed to ambient pressures. The two chambers are separated by a membrane. The valve actuator assembly is also operationally associated with a fluid pressure intensifier that boosts the ambient pressure of the accumulator so that an increased fluid pressure may be transmitted to the actuator assembly to bias the actuated valve toward its fail safe configuration. In a described embodiment, the fluid pressure intensifier comprises a housing that defines a chamber having a fluid inlet and fluid outlet. A dual-headed piston is moveably retained within the housing. The piston has an enlarged piston face and a reduced size piston face. Fluid pressure entering the fluid inlet is exerted upon the enlarged piston face, and due to the difference of piston face sizes, an increased pressure is transmitted out of the fluid outlet.

A face importance level determining apparatus for determining an importance level for a photographer with respect to a face of a person appearing in an image may include the following elements. A face information obtaining unit obtains face size information and position information that are detected from an input image signal. A first calculating unit calculates a first importance level factor on the basis of the size information. A second calculating unit calculates a second importance level factor on the basis of the position information. A third calculating unit calculates the importance level on the basis of the first importance level factor and the second importance level factor.

A data processing device includes: a face-detecting unit to detect a face region from an input image input from an imaging unit which images a subject including a face; a setting unit to calculate subject distance based on the face size of the face detected by the face-detecting unit, and set limits as a focus lens operation range before and after a focus lens position with the calculated subject distance as focal distance; a detecting unit to move a focus lens within the limits set by the setting unit, and detect a focal point where a detection value corresponding to the contrast intensity obtained from a face frame including the face detected by the face-detecting unit is not smaller than a threshold; and a control unit to determine, in a case wherein a focal point has been detected within predetermined set time, the focal point thereof as a focus position.

The present invention discloses an identification photo camera capable of performing human image posture photography prompting and a human image posture detection method. The identification photo camera comprises a human image feature acquisition module, a facial feature detection module, a human face size and position detection module, a human image photography distance detection module, a human face state detection module and a shoulder state detection module. The method comprises: acquiring human image feature data of a preview frame image; and according to the acquired human image feature data, judging whether facial features are complete or not, the size and position of the human face meet requirements or not, a distance from a photographed person to a camera head is an optimal photography distance range or not, the human face is regular or not, and the shoulders are flush or not. According to the identification photo camera and the method, professional human image posture photography prompting can be provided for a user, and operation is performed according to the photography prompting, so that the photography quality of an identification photo is greatly improved.

The invention discloses a real-time face optimal selection method based on a video sequence. The method comprises the following steps: a face image is acquired; face clarity, face size and opening degree of human eyes act as three indexes for face quality evaluation so that a face integrated evaluation score ImageScore is calculated; the ImageScore of the first frame of the face image is assigned to an initial face optimal image MaxScore and the face image is stored; the MaxScore after assignment is compared with the ImageScore of other frames of the face image, and the face image corresponding to the higher ImageScore is stored and the ImageScore is assigned to the MaxScore again; and the MaxScore value is cyclically updated and the stored optimal face image is updated until the image sequence with the face cannot be acquired, and the cycle is ended and the image corresponding to the stored MaxScore is the optimal face image. The face image with higher quality can be screened for identification via comparison of the integrated evaluation indexes of face quality in the video sequence under the situation of not knowing the face sequence image in advance; besides, the calculation method is simple with real-time performance.

A face recognition apparatus and method. Sub-images having different face sizes are generated using a received face image of a person to be identified. Feature vectors of the sub-images are generated and observation nodes are generated based on the feature vectors. The observation nodes corresponding to the sub-images are compared with stored reference nodes of sub-images of a registered person on a face size by face size basis to calculate similarity scores between the observation nodes and the reference nodes. State nodes are generated based on the respective similarity scores of the face sizes, the observation and state nodes are compared, and the state nodes are compared to perform face recognition. This improves face recognition performance and face recognition speed. Face recognition performance robust to facial expression variation or type information is achieved by performing I-shaped curvature Gabor filtering on a plurality of sub-images based on the eye distance.

Autofocus operations in a digital camera are described. In one embodiment, a processor detects a face in an image captured by the camera, and calculates the size of the face. It selects from amongst a number of previously stored face sizes, one that is closest to the calculated face size. It retrieves a previously stored lens focus position that is associated with the selected, previously stored face size. It signals a moveable lens system of the digital camera to move to a final focus position given by the retrieved, previously stored lens focus position. Other embodiments are also described and claimed.

A face detection apparatus associated methodology is provided to generate an output indicating a likelihood of a region of an image containing a face. The apparatus includes a comparing device for comparing the region with face data that indicates a presence of the face and for detecting respective likelihood values indicating the likelihood of the region containing the face in a group of respective different face size. A control device is provided for modifying at least one of the group and the likelihood values based on a face size or a range of face sizes appropriate to at least one classification of the image. The control device is responsive to metadata associated with the image that define the at least one classification of the image. The image is a part of a video sequence, and at least one classification includes a video program type.

The invention discloses a face detection method and system based on the flexible convolution network. The method comprises steps that an input image is received, and the image includes the to-be-detected face information; features of the image are extracted through the main network, and the features include semantic features and offset features; the semantic features and the offset features of theimage are respectively outputted through flexible convolution; according to the semantic features of the image, the face containing probability of an anchor box is outputted through the classification sub network; according to the semantic features of the image, the overlapping degree IOU between each prediction frame and a face is outputted through an IOU prediction branch; according to the offset features of the image, the face size and the position corresponding to each prediction frame are outputted through the position regression sub network. The method is advantaged in that the face inthe image can be accurately predicted.

In a digital camera, in-focus position data representative of a distance from a taking lens to a focused subject is obtained from a detection result of an auto-focus detection circuit. Also angle-of-view data representative of an image capturing field is obtained from record of operation on a zooming button. Based on the in-focus position data and the angle-of-view data, a face size determination circuit determines the size of a face that may be contained in a captured image. A face detection circuit detects a face using a frame whose size is defined by the determined face size. A detection result of the face detection circuit is displayed with the target image on an LCD panel. If required, a data base is revised to reflect the detection result.

The invention discloses an eyesight protection system and a method based on three-dimensional videography, wherein the method comprises the following steps that: a display is taken as a videography center to obtain a human body image positioned in front of the display, and the position of a face setting zone is detected inside the human body image; according to the obtained position information of the face setting zone, the image of the face setting zone is determined inside the human body image; according to the image of the face setting zone and the calibration information of a three-dimensional videography unit, the distance between eyes and a display screen is calculated; moreover, when the calculated distance is less than a set protection distance, warning information is sent out. The technical proposal disclosed by the invention does not need to determine a safety distance according to the statistical information of face size, and eliminates the influence of face size on a calculated result.

An image pickup device can effectively detect a face in accordance with an object without causing a user to do bothersome setting. A speed preference mode and a person-number preference mode both having different detection conditions of a face region are prepared beforehand as face detection modes. The user is caused to select a desired face detection mode, and minimum values of a face size set as detectable in a face detection and a maximum number of face regions to be detected are set to values corresponding to the face detection mode selected by the user. In a case where an image-pickup preparation is set by pressing a shutter key halfway, when the face detection mode is the speed preference mode, the mode is forcibly changed to the person-number preference mode.

The invention provides a pedestrian position detection and tracking method and system. The pedestrian position detection and tracking method comprises the following steps: acquiring a video stream ina preset scene, sequentially inputting frame-by-frame images in the video stream into a face detection model, and outputting a face position and a face size target frame of a pedestrian in the video stream; estimating a pedestrian target frame according to the body characteristics of the pedestrian and the face size target frame, fusing the estimated pedestrian target frame with the result of thepedestrian target frame detected by the background subtraction method, and performing correction to obtain a corrected pedestrian target frame; obtaining a position corresponding to the corrected pedestrian target frame according to a mapping relationship between the pedestrian target object and a preset scene geographic position; and calculating the centroid of the corrected pedestrian target frame to serve as the centroid of the pedestrian target of the corresponding frame image, and tracking the centroid of the pedestrian target of each frame image in the preprocessed video stream to realize pedestrian tracking. The method has good detection and tracking effects, and the influence caused by multi-target overlapping is avoided.

The invention relates to the technical field of computer vision and virtual wearing combination and discloses a virtual glasses trying method based on a mobile terminal. The virtual glasses trying method specifically comprises the following steps of: (1) calculating the face size of a user according to the known size of an identification object; (2) taking down the identification object by the user, obtaining a real-time face image by a camera, marking face feature points in the real-time face image by adopting a fast face recognition algorithm, estimating head attitude information and face shape information according to marking results, and determining position coordinates of a pair of virtual glasses according to the face size obtained in the step (1); (3) obtaining selected glasses information, so as to obtain a glasses image, overlapping the image on the coordinates determined in the step (2), and then displaying the face image overlapped with the glasses on a display screen of the mobile terminal in real time, so that the user conveniently has good shopping experience. The invention further discloses a virtual glasses trying system based on the mobile terminal.

The invention relates to a symmetrical composite roll forming method of an external step ring with a thick wall and a small hole, which is characterized by comprising the following steps of: (1) blanking: determining the size of a roll-formed external groove ring according to the size of the external step ring with the thick wall and the small hole, and then determining the size of a ring blank based on the principle that the volume of plastic deformation is unchanged; (2) design of grooves: designing the working face sizes of a king roller and a core roller according to the rolling deformation conditions of the ring and the structural requirements of equipment, and then determining the working face size and position of an auxiliary roller according to the sizes of the ring blank, the ring and the king roller; (3) design of forming parameters: designing the rotating speed of the king roller and the feeding speed of the king roller in sequence according to the rolling deformation conditions of the ring and the size of the king roller; (4) roll forming; and (5) cutting. Through the method, the external step ring with the thick wall and the small hole, such as a pipeline connecting flange, a high pressure valve body and the like in petrochemical and power fields, can be manufactured with low consumption, high efficiency and high performance.

The invention discloses a single person close-up real-time recognition and automatic screenshot method for a large live scene. The method particularly comprises the following steps: a current video frame image in a live video is acquired in real time, and whether a face exists in the video frame image is detected; if a face exists in the video frame image, a face detection module is adopted to acquire a face area, comprehensive evaluation of a face size, definition, a location and an angle is carried out on the face area, an optimal face is further selected, and the current video frame image containing the optical face is subjected to screenshot saving; and finally, the optimal face image is fed to a face recognition model for recognition, and a recognition result is outputted. The methodcan be applied to a large live scene for recognition and screenshot on a single person close-up in the video, automatic screenshot saving is carried out when a screenshot index is met, a screenshot with a better quality can be obtained due to the screenshot index, and the problem of large calculation workload of a backend server brought by a large amount of repeated face snapshots generated in face recognition can be solved.

The invention provides a camera automatic variable-time control method based on the human eye space. According to the method, a zoom lens and / or zoom driving motor or a prime lens, an image sensor and a video analysis processing unit are mainly used, the front side of a person enters a shooting view field, the two eyes in the video are locked, the space w of the two eyes is measured and calculated, at the moment, the lens focus f and the human eye general space W and the human eye space w in the video are known quantities, the values are substituted into a formula, the distance of an object to be shot can be calculated, the expected human face size is set in a monitoring image, the space between the two eyes in the image is wi, the space w of the two eyes in an actual image is measured and calculated, the ratio of the w to the wi is calculated, the zoom power of the lens is regulated according to a preset relational expression in accordance to the proportion relationship, so that the regulated w equals to the wi, the steps are repeated when the distance between the human face and the lens is changed, and the w is enabled to be equal to the wi, so that the human face can be enabled to be always displayed in the image in the preset size in a certain distance range.

A gain adjusting system for adjusting a gain of a sound signal in an audio system, includes a first detecting unit for capturing images of one or more faces of users and determining the number of faces and the size of the faces present in the images; a controller for receiving face data from the first detecting unit for comparing the sizes of faces in subsequently captured images with an initial face size and accordingly deciding and outputting a first decision signal; and a gain regulator coupled to the controller for adjusting the gain level of the sound signal according to the first decision signal.

An image processing apparatus for processing an image including an image of a face includes: a normalization unit configured to normalize the image including the image of a face so that the size of the face becomes a predetermined face size; a detection area setting unit configured to set an area smaller than the image normalized by the normalization unit as a detection area in which a position of a face part of the face is detected; and a detection unit configured to detect the position of the face part in the detection area set by the detection area setting unit.

A method for photographing an image in a device having a camera according to the present invention includes displaying a preview image, identifying a face size of a person included in the preview image, displaying a guide in the preview image, according to a photographing type determined by the identified face size, and photographing and storing an image by capturing the image output by the camera when photographing is requested.

An eyebrow guide configured as an adjustable facial frame designed to be centered over the nose and having a pair of selectively shaped and positioned eyebrow templates for use in guiding the appliance used to shape, color, shave, cut or otherwise cosmetically treat the user's eyebrows. The frame comprises a nosepiece designed to overly the nose to both center and stabilize the guide. The frame also has a pair of opposed side members which extend from the nosepiece toward and over the ears in a manner similar to conventional eyeglass frames to further stabilize and support the guide. Intermediate of the eyebrow templates and the nose piece is a three-piece, double hinged member designed to provide adjustment for virtually any facial dimensions in regard to distance between the eyebrows and the nose as well as their relative depth. In addition, the upper portion of this double-hinged member provides a dial for eyebrow angle adjustment for the templates.

The embodiment of the invention provides a shooting method which is applied to a mobile terminal with a camera. The shooting method particularly can comprise the steps of: acquiring one frame of preview image acquired by the camera of the mobile terminal; acquiring a shooting distance and a face size value of each face in the preview image; based on the shooting distance and the face size value of each face, determining at least one target face to be amplified in the preview image; carrying out amplification processing on the at least one target face to generate a first target image; and generating a second target image from the preview image, and carrying out association storage on the first target image and the second target image. According to the shooting method provided by the embodiment of the invention, amplification on a face at a long shooting distance can be implemented, so that features of five sense organs of a target figure are clearly displayed, and an imaging effect of the image is improved; and moreover, the second target image which is not processed at all and the first target image subjected to amplification processing can be provided for a user to select, so that selections of the user are enriched.

The invention discloses a multi-person close-up real-time recognition and automatic screenshot method for a large live broadcasting scene. The method particularly comprises steps: a face size, definition and a face position are used as first, second and third indexes for a screenshot; firstly, the current video frame image of a live broadcasting video is acquired in real time, a face in the videoframe image is detected, and a face area is acquired and tracked; and then, the face size, the definition and the position in the face area are evaluated sequentially, a face image meeting a screenshot standard is further selected, the current video frame image screenshot is saved and sent to a face recognition model for recognition, and a recognition result is outputted. The method can be appliedto a large live broadcasting scene to carry out recognition and screenshot on multiple person close-ups in the video, automatic screenshot for saving is carried out when a screenshot index is met, and a screenshot with a better quality is ensured to be acquired.