159 results about "Face analysis" patented technology

A face feature analysis which begins by generating multiple face feature candidates, e.g., eyes and nose positions, using an isolated frame face analysis. Then, a nostril tracking window is defined around a nose candidate and tests are applied to the pixels therein based on percentages of skin color area pixels and nostril area pixels to determine whether the nose candidate represents an actual nose. Once actual nostrils are identified, size, separation and contiguity of the actual nostrils is determined by projecting the nostril pixels within the nostril tracking window. A mouth window is defined around the mouth region and mouth detail analysis is then applied to the pixels within the mouth window to identify inner mouth and teeth pixels and therefrom generate an inner mouth contour. The nostril position and inner mouth contour are used to generate a synthetic model head. A direct comparison is made between the inner mouth contour generated and that of a synthetic model head and the synthetic model head is adjusted accordingly. Vector quantization algorithms may be used to develop a codebook of face model parameters to improve processing efficiency. The face feature analysis is suitable regardless of noise, illumination variations, head tilt, scale variations and nostril shape.

A processor-based system operating according to digitally-embedded programming instructions performs a method including identifying a group of pixels corresponding to a face region within digital image data acquired by an image acquisition device. A set of face analysis parameter values is extacted from said face region, including a faceprint associated with the face region. First and second reference faceprints are determined for a person using reference images captured respectively in predetermined face-portrait conditions and using ambient conditions. The faceprints are analyzed to determine a baseline faceprint and a range of variability from the baseline associated with the person. Results of the analyzing are stored and used in subsequent recognition of the person in a subsequent image acquired under ambient conditions.

An apparatus and method for image-based 3D photorealistic head modeling are provided. The method for creating a 3D photorealistic head model includes: detecting frontal and profile features in input frontal and profile images; generating a 3D head model by fitting a 3D genetic model using the detected facial features; generating a realistic texture from the input frontal and profile images; and mapping the texture onto the 3D head model. In the apparatus and method, data obtained using a relatively cheap device, such as a digital camera, can be processed in an automated manner, and satisfactory results can be obtained even from imperfect input data. In other words, facial features can be extracted in an automated manner, and a robust “human-quality” face analysis algorithm is used.

The invention relates to a face recognition system and method. The system at least comprises a data input module, a face analysis module and a data output module; the data input module provides an image sequence to be detected for the face analysis module; with the help of a deep learning technology, the face analysis module performs face recognition, obtains similar faces through a KD tree, performs parallel retrieval, comparison and analysis based on a plurality of sub-databases, and combines analysis results; and the data output module overlaps a face detection frame, face relevant statistical information and original videos and outputs to the front end to display. The invention further provides a method for realizing the system. The system disclosed by the invention has the characteristics of being automatic, rapid and precise while recognizing faces.

The invention provides a teaching quality evaluation system based on human face analysis, which belongs to the technical field of intelligent detection and analysis and solves the technical problem of modeling analysis between the teaching quality and classroom performance of a student. According to the teaching quality evaluation system based on human face analysis, the following several parts are included: S1, a human face recognition attendance module, S2, human face posture analysis, S3, facial expression analysis, S4, building of the relationship between the classroom performance and the evaluated teaching quality based on big data analysis, S4, a teaching quality evaluation method based on classroom performance, and S6, a teaching quality feedback mechanism based on big data analysis. The teaching quality evaluation system based on human face analysis has the advantages that through modeling analysis, the teaching quality is improved, and the teaching quality evaluation can be quantized.

An image processing system is disclosed that provides automatic face or skin blurring for images. All faces or skin can be blurred, or specific faces can be blurred. In one aspect of the invention, a particular face is blurred on an image or on a series of images in a video. Faces are determined in an image, and face matching is performed to match a particular face to faces in the image. If a match is found, the face or a portion of the face is blurred in the image. The blurring is performed on a portion of the image containing the particular face. Blurring may be performed through a variety of techniques. In another aspect of the invention, voice processing is used as an adjunct to or in place of face analysis to determine if a face in an image or series of images should be blurred. In another aspect of the invention, all faces or human skin in an image or series of images is blurred.

The invention provides a CNN based face image quality evaluation method and device. The method comprises that face images in a monitoring video are collected to form a face data set, and the face dataset comprises an initial quality value corresponding to each face image; image data of each face image is calculated, a training sample set is generated according to the image data and initial quality value of each image, and the image data includes a binary key point image, a grayscale map and an edge intensity graph; the training sample set is trained on the basis of the CNN, and a quality evaluation model is obtained; and a to-be-identified face image generates image data, the face quality of the image data is evaluated by the quality evaluation model, and a quality value of the to-be-identified face image is obtained. Via the schemes, a frame with the highest face image quality in the video can be used for subsequent face analysis and identification. The quality evaluation and filtering effect can be optimized, and the case solving efficiency can be improved greatly.

An apparatus and method for image-based three-dimensional hyperrealistic head modeling are provided. The method used to create a three-dimensional hyperrealistic head model includes: detecting frontal and side features in the input frontal and side images; generating a three-dimensional head model by fitting a three-dimensional universal model using the detected facial features; according to the input generating realistic textures from front and side images; and mapping said textures onto a 3D head model. In the described apparatus and method, data obtained using relatively inexpensive equipment such as digital cameras can be processed in an automated manner, and satisfactory results can be obtained even from imperfect input data. In other words, facial features can be extracted in an automated manner and using stable "human quality" facial analysis algorithms.

The invention discloses a face analysis method and a face analysis system integrating a multi-task combined multi-scale convolution neural network, Firstly, the key region search algorithm is used toextract K face regions of interest with different scales from a picture of N *N size, which is used as the input of three channels of multi-scale CNN. Then, CNN is used to extract the features of theK regions of interest to obtain the features of faces with different scales, and the extracted features of faces with different scales are fused in a cascade manner to obtain the fused feature expression; Then, the loss functions of a plurality of tasks are fused to obtain a joint loss function, and the feature expression is used as a learning input to obtain an optimal solution of the joint lossfunction, so as to obtain a processing result of the plurality of tasks. The invention utilizes the correlation between tasks to promote each other, and improves the prediction accuracy rate of a single task.

This invention relates to an 2DFDA man-face identification method based on block relativity, which divides a man-face image into non-overlapped small blocks according to the regianl characters of a man-face image, then connects the elements in each small block in lines to generate related line vectors and arranges the vectors in sequence to a 2-D image matrix to be taken as the input image finally for 2DFDA identification, which fully utilizes the relativity information between line and row pixels in local regions and remains the local character information to get rather high identification rate of man faces.

The invention discloses a method, a device and a system for generating a cartoon face picture. The method comprises the following steps: transmitting a face picture to a face analysis server, wherein the face analysis server is used for extracting face features and character attribute according to the face picture; receiving the face feature and character attribute from the face analysis server, searching cartoon materials matched with the face features and character attribute from a cartoon material library; generating a cartoon face picture according to the cartoon material. By adopting the method, the business processing workload of a terminal can be alleviated, and the operation speed of the terminal can be increased.

The invention discloses a face image reversible geometric normalization method facing overall characteristics analysis. A method based on secondary affine transformation is adopted. Firstly global affine is carried out which means that global rotation, zooming and cutting are carried out on a face image according to a standard structure definition of a face, such that in an input face image, coordinates of two eyes are at a same horizontal line, a two eye center distance is subjected to normalization according to a parameter d of a standard structure, and transverse and longitudinal proportions of the face are consistent with the standard structure. The image is cut into a magnitude of 2d*2d. Then geometric position normalization is carried out on a mouth area through a local affine transformation needle. Finally by utilizing image edge fusion, a standard geometric normalization face image is formed. An inverse normalization process is that according to a transformation parameter, a standard structure face image is recovered to a geometric structure characteristic before geometric normalization through secondary affine inverse transformation such that personalized topology structures of different face characteristic points are maintained. The face image which goes through geometric normalization can be used for face analysis based on an overall characteristic, such as super resolution reconstruction of the face image, face recognition, and the like.

The invention provides a human face emotion analysis method and system based on multi-task learning and deep learning. The method comprises the steps of learning preset analysis task convolution layers in a human face database by using a convolutional neural network to obtain a human face analysis model; acquiring a to-be-analyzed human face image, analyzing the to-be-analyzed human face image by using a human face detection algorithm and extracting human face areas in the to-be-analyzed human face image; performing prediction on the to-be-analyzed human face image by using the human face analysis model and obtaining emotion information corresponding to each human face area in the to-be-analyzed human face image. According to the invention, the concept of multi-task learning is applied to a convolutional neural network and multiple kinds of analysis tasks related to human faces can be identified by using the same analysis model, so that the size of an analysis model is reduced and identification is accelerated; different parts of a human face are described by using different convolution layers, so that each convolution layer focuses on one single task and a better identification effect can be achieved.

Disclosed is an infrared human face spectacle disturbance elimination method based on regional characteristic element compensation, proceeding eyeglass detection to the imported infrared human face graph, carrying out corresponding treatment based on whether eyeglass is present, for infrared human face graph with no eyeglass, carrying out characteristic face analysis for direct neighboring categorization, thus obtaining the discrimination result, for infrared human face graph with eyeglass, carrying out characteristic face analysis to obtain characteristic vector, compensating the provincial characteristics elements of the eyeglass, proceeding neighboring categorization to obtain the discrimination result. The invention can effectively repress the disturbance by eyeglass.

The invention discloses an engine driver state monitoring method and a device based on video face analysis. The method comprises the following steps: collecting the monitoring video images of an engine driver in real time; conducting face detection to the collected monitoring video images to obtain the detecting results of acquired driver face images or driver face images which are not acquired; analyzing the detecting results and acquiring the information of the on-duty state of the driver; according to the information of the on-duty state of the driver, confirming if the driver leaves a post, and issuing reminding and alarming information of the driver when confirming that the driver leaves the post; acquiring the fatigue state information of the driver through the analysis of the acquired face image of the driver; and according to the fatigue state information of the driver, confirming if the driver is fatigue, and when the driver is confirmed to be fatigue, issuing fatigue reminding and alarming information of the driver. According to the engine driver state monitoring method and the device based on the video face analysis, automatic identification and alarming reminding of the engine driver state can be completed, and automatic identification alarming and real-time scene reminding of the engine driver state are achieved.

The invention discloses a public bus passenger flow volume detection system and method based on face identification and position positioning. The Public transportation passenger flow volume detection system comprises an image collection module, a public bus position collection module, a data mining center and data analysis center which are successively connected; the image collection is used for collecting human faces of people getting on and off the public bus; the public bus position collection module is used for recording the driving course of the public bus; the data mining center is used for performing human face analysis and position analysis on the collected image; the data analysis center is used for performing flow statistic analysis according to the result obtained by the data mining center and providing the decision basis to the public bus route planning, operation management, route adjustment, etc. The pblic transportation passenger flow volume detection system and method not only improve the public bus passenger flow volume accuracy, improvethe public bus passenger flow volume detection frequency and greatly saves the flow investigation cost.

The invention implements a traditional Chinese medicine intelligent diagnosis expert system, which comprises an image acquisition sub-system, an image processing sub-system, a pulse condition detection sub-system, a pulse condition processing sub-system, a comprehensive analysis sub-system and a database management sub-system, wherein the image acquisition sub-system and the image processing sub-system are coordinated in work, so that tongue diagnosis and face diagnosis detection functions are achieved, and tongue / face analysis reports are generated; the pulse condition detection sub-system and the pulse condition processing sub-system are coordinated in work, so that a pulse diagnosis detection function is achieved, and a pulse condition analysis report is generated; and finally, the comprehensive analysis sub-system is in charge of summarizing the two analysis reports so as to offer a diagnosis result and treatment suggestions, and the diagnosis result and the treatment suggestions are fed back to users. In addition, the system further comprises the database management sub-system which is in charge of managing personal data of all registered users. The traditional Chinese medicine diagnosis system provided by the invention has the remarkable advantages that traditional Chinese medicine tongue diagnosis, face diagnosis and pulse diagnosis are combined, so that a comprehensive and detailed diagnosis result can be obtained; and traditional Chinese medicine is promoted.

The invention provides a similarity discrimination system based on the five sense organs of a human face. An initialization module is used for establishing an initial human face image model. A five-sense-organ positioning module is used for respectively acquiring the positions of five sense organs in the human face region of a plurality of to-be-detected human face images, dividing and extracting the region of each sense organ in the human face region. A feature extraction module is used for obtaining the similarity information of each sense organ in the human face region of the to-be-detected human face images so as to form five-sense-organ feature vectors. A judging module is used for comparing the five-sense-organ feature vectors of a to-be-detected human face with five-sense-organ feature vectors of already known human faces so as to obtain the similarity of the to-be-detected human face. Based on the system, the face features of people are basically classified through judging the similarity of the five sense organs of a human face. Meanwhile, the accurate and detailed analysis and comparison for face features is conducted. As a result, the similarity degree between the sense organs of different people or between the sense organs of people in different images can be provided for users. Furthermore, the system comprehensively embodies the similarity degree between human faces, and improves the human face analysis accuracy and the human face analysis comprehensiveness.

The invention relates to a complex makeup migration method based on facial analysis and illumination migration. Facial analysis is combined with an illumination migration algorithm and a makeup migration algorithm. The method comprises the following steps: firstly, carrying out whitening and buffing preprocessing on an input face image by using a color balance and bilateral filtering algorithm; secondly, migrating makeup of the reference face image to the input face image by utilizing a face analysis and facial makeup migration algorithm, and finally migrating illumination of the reference face image to the input face image by utilizing an illumination migration algorithm to complete migration of complex facial makeup. The method is very easy to realize through software, and can be widelyapplied and popularized to computer vision, visual design, digital entertainment, film production and the like.

The invention discloses a passenger flow detection system based on face recognition and a detection method thereof. The key points of the technical scheme are that the passenger flow detection systemcomprises an image acquisition module and a passenger flow statistical server. The passenger flow statistical server comprises a storage module, a face analysis module and a flow analysis module whichare connected in turn. The image acquisition module is installed at the entrance and used for acquiring various angles of face image information of the target person. The storage module is used for storing the acquired image information of the target person. The face analysis module is used for recognizing the face information of each image and performing comparative analysis with other images inthe storage module. The flow analysis module is used for integrating the analysis result outputted by the face analysis module so as to obtain the passenger flow characteristics through statistics and form the trend chart. According to the passenger flow detection system based on face recognition, the customer information can be more deeply read and the accuracy can be greatly enhanced.

The invention discloses a passenger flow analysis system based on face recognition and a deployment method. The passenger flow analysis system comprises a user information module, a video monitoring module, a face recognition module, a face analysis module, a passenger flow statistics module and a passenger flow prediction module. After the system is applied, the system can accurately identify genders and age groups of customers and accurately subdivide customer groups; the system can intelligently predict the number of users, age and gender composition and people flow peak time in any time period according to historical data, and the management and operation efficiency is improved; the system adopts a B / S architecture, monitoring videos and statistical analysis data can be checked throughpersonal terminals such as a mobile phone and a PC, and data mobility management is realized; and all data are stored and analyzed on the server, so that the space of personal equipment is not occupied, and the data analysis rate and the data security are ensured.

The invention relates to a body temperature measuring method based on infrared image face detection. The body temperature measuring method comprises: obtaining an infrared image and a temperature distribution diagram through an infrared temperature measuring camera; performing face analysis based on the infrared image to obtain a face area; based on the temperature distribution diagram, performingtemperature analysis, and determining a body temperature area; calibrating the face region according to the body temperature region, and determining a region of interest; and calculating the averagetemperature of the region of interest according to a preset algorithm. Remote portable body temperature measurement can be realized, the face area of the measured human body is accurately positioned through face analysis of the infrared image, the face area is calibrated in combination with the temperature distribution diagram, and the method has the advantages of high measurement accuracy and lowomission ratio.

The invention relates to the technical field of computer graphics, and particularly relates to an expression editing method and device based on face analysis. The purpose of the invention is to solvethe problem of how to keep figure identity and carryin out expression editing on the expression-free face. In order to achieve the purpose, the expression editing method based on face analysis provided by the invention comprises the following steps: obtaining an expression-containing face analysis image corresponding to an expression-free face image and an expression intensity control code based on a preset face analysis prediction network and according to the expression-free face image and the expression intensity control code which are obtained in advance; And obtaining a target expression image based on a preset face expression editing network and according to the expression-containing face analysis image and the expression-free face image. Based on the above steps, editing of differentintensities and different expressions can be carried out on the face without expressions on the premise of keeping the identity information of the character.

The invention utilizes a face analysis and recognition technology to propose a similarity analysis method, device and system used for a celebrity matching game. First, a HypeFace method is utilized toperform face detection, key point positioning, and posture and gender analysis on an input original image. Next, a machine vision technology is utilized to perform resolution ratio, posture, expression and sheltering analysis on a detected face to judge qualification. Then, a face recognition technology is utilized to perform celebrity matching on the detected face, and finding out a celebrity which is the most similar to the input face. To improve stability of celebrity matching, a method of combining overall facial form matching and local organ matching is adopted. Finally, a matching result is displayed, including facial form similarity, similarity of organs, the most similar organ, the most dissimilar organ and brief introduction of the celebrity. The abovementioned method rapidly andaccurately finds out the most similar celebrity through the face analysis and recognition technology, and performs comprehensive analysis and display of similarity, thereby improving user experience.

An image transformation apparatus includes a detection unit which is configured to detect, from each of a user image and a reference image, feature points of a face and angle information of the face, a feature points adjusting unit which is configured to adjust the feature points of the user image or the reference image by using the detected angle information, a face analysis unit which is configured to compare facial features contained in the user image and the reference image by using the adjusted feature points, and an image transformation unit which is configured to transform the user image by using a result of the comparison of the facial features from the face analysis unit.

A deep learning framework jointly optimizes the compactness and discriminative ability of face representations. The compact representation can be as compact as 32 bits and still produce highly discriminative performance. In another aspect, based on the extreme compactness, traditional face analysis tasks (e.g. gender analysis) can be effectively solved by a Look-Up-Table approach given a large-scale face data set.