National Yang Ming Chiao Tung University

Feiya Technology Corp.

Silicon Motion, Inc.

Faceheart Inc.

Cssp Inc.

Bing-Fei Wu

Meng-Liang Chung

Po-Wei Huang

Chao-Jung Chen

Chun-Hsien Lin

Yen-Lin Chen

Chih-Chung Kao

Yi-Chiao Wu

Cheng-Chih Chien

Bing-Jhang Wu

Hao-Yu Huang

Kuan-Hung Chen

Chung-Cheng Chiu

Min-Yu Ku

Tzu-Min Lin

Chih-Chun Liu

Yin-Cheng Tsai

Cheng-Yen Yang

Yun-Wei Chu

Ying-Han Chen

Jou-Wei Fu

Wen-Chung Chang

Chun-Hao Lin

Yu-Ming Chen

Khein Seng Pua

Chih-Cheng Huang

Hsin-Yu Lin

Wen-Hsin Lin

Chih-Wei Liu

Lung-Yi Kuo

Chung-Chi Tien

Aw Yong-Chee Kong

Wei-Hsiang Su

Chung-Jui Fan

Tsong-Yang Tsou

Yin-Yin Yang

Hsin-Yuan Peng

Wan-Ju Tseng

Yuan-Hsin Chen

A system of image processing based emotion recognition is disclosed. The system principally comprises a camera and a main processor. Particularly, there a plurality of function units provided in the main processor, including: face detection unit, feature processing module, feature combination unit, conversion module, facial action judging unit, and emotion recognition unit. According to the present invention, the emotion recognition unit is configured to utilize a facial emotion recognition (FER) model to evaluate or distinguish an emotion state of a user based on at least one facial action, at least one emotional dimension, and a plurality of emotional scores. As a result, the accuracy of the emotion recognition conducted by the emotion recognition unit is significantly enhanced because basis of the emotion recognition comprises basic emotions, emotional dimension(s) and the user's facial action.

System and method of image processing based emotion recognition

A device for liveness detection is disclosed. The liveness detecting device has a simplest structure that principally comprises a light sensing unit and a signal processing module. Particularly, the signal processing module is configured for having a physiological feature extracting unit and a liveness detecting unit therein. The physiological feature extracting unit is adopted for extracting a first physiological feature from a PPG signal, or extracting a second physiological feature from the PPG signal that has been applied with a signal process. As such, through the first and second physiological features, the liveness detecting unit is able to determine whether a subject is a living body or not. The liveness detecting device does not use any camera unit and iPPG technology, such that the liveness detecting device has advantages of simple structure, low cost and immediately completing liveness detection.

Device and method for liveness detection

A method for assessing driver fatigue is implemented by a processor and includes steps of: based on images of a driver captured by an image capturing device, obtaining an entry of physiological information that indicates a physiological state of the driver; based on one of the images of the driver, obtaining an entry of facial expression information that indicates an emotional state of the driver; based on one of the images of the driver, obtaining an entry of behavioral information that indicates driver behavior of the driver; and based on the entry of physiological information, the entry of facial expression information and the entry of behavioral information, obtaining a fatigue score that indicates a level of fatigue of the driver.

Method for assessing driver fatigue

The present invention provides a physiological information detection method for calculating a physiological value by using changes of a dynamic image. The detection method includes: acquiring detection data from a gray-scale value of the dynamic image, and transforming the detection data into frequency data. The detection method further includes: determining whether the frequency data meet a preset condition, and using a transformation model of a corresponding transformation combination accordingly to transform the frequency data into a physiological value. The present invention further provides a physiological information detection device applying the detection method.

Physiological information detection device and physiological information detection method

A people-flow analysis system includes an image source, a computing device, and a host. The image source captures a first image and a second image. The computing device is connected to the image source. The computing device identifies the first image according to a data set to generate a first detecting image. The first detecting image has a position box corresponding to a pedestrian in the first image. The computing device generates a tracking image according to the data set and a difference between the first detecting image and the second image. The tracking image has another position box corresponding to a pedestrian in the second image. The host is connected to the computing device and generates a people-flow list according to the first detecting image and the tracking image.

People-flow analysis system and people-flow analysis method

A non-contact heartbeat rate measurement system includes an image sensor, a target region selecting module, a heartbeat signal calculating module, a spectrum analyzing module, a vibration detecting module, a heartbeat peak selecting module. When a signal quality indicator is greater than a threshold value, and a first peak with global highest signal intensity in a heartbeat spectrum is similar to a face vibration frequency, the heartbeat peak selecting module selects a second peak with local highest signal intensity from a part frequency band of the heartbeat spectrum as an output heartbeat frequency. A non-contact heartbeat rate measurement method and an apparatus are disclosed herein.

Non-contact heartbeat rate measurement system, method and apparatus thereof

A recycling system and a method based on deep-learning and computer vision technology are disclosed. The system includes a trash sorting device and a trash sorting algorithm. The trash sorting device includes a trash arraying mechanism, trash sensors, a trash transfer mechanism and a controller. The trash arraying mechanism is configured to process trash in a batch manner. The controller drives the trash arraying mechanism according to the signals of trash sensors and controls the sorting gates of the trash sorting mechanism to rotate. The trash sorting algorithm makes use of the images of trash, wherein the images are taken by cameras in different directions. The trash sorting algorithm includes a dynamic object detection algorithm, an image pre-processing algorithm, an identification module and a voting and selecting algorithm. The identification module is based on the convolutional neural networks (CNNs) and may at least identify four kinds of trash.

Recycling system and method based on deep-learning and computer vision technology

The invention is related to a method of three-dimensional face reconstruction by inputting a single face image to reconstruct a three-dimensional face model, therefore, the human face image is seen at various angles of three-dimensional face through rotating the model images.

Three-dimensional human face reconstruction method

The present invention provides a contactless-type sport training monitor method, comprising: selecting at least an image database to recognize a plurality of expressions in the image database; making pre-processing for the plurality of expressions: using a convolutional neural network as a feature point extraction model; acquiring a human image; tracking a first target region and a second target region in the human image; making chrominance-based rPPG trace extraction; using the deep level model to compare the second target region image; and to calculate a post-exercise heart rate recovery achievement ratio, to judge the Rating of Perceived Exertion, and judging whether the human body is under overtraining status or not.

Contactless-type sport training monitor method

A facial expression recognition training system includes a training module, feature database, a capturing module, a recognition module and an adjusting module. The training module trains a facial expression feature capturing model according to known face images. The feature database stores known facial expression features of the known face images. The capturing module continuously captures first face images, and the facial expression feature capturing model outputs facial expression features of the first face images according to the first face images. The recognition module compares the facial expression features and the known facial expression features, and fit the facial expression features to the first known facial expression features that is one kind of the known facial expression feature accordingly. The adjusting module adjusts the facial expression feature capturing model to reduce the differences between the facial expression features and the known facial expression features.

Facial expression recognition training system and facial expression recognition training method

The present invention provides an image based blood pressure monitoring method, comprising: acquiring at least a human image information of at least a human skin area; according to the human image information to locate at least a ROI; extracting the human image information of the ROI, and calculate; filtering the average value of the human image information; monitoring the filtered signal; calculating an image pulse transmit time of the filtered signal, and calculating an inter-beat interval of the filtered signal; and employ the specific prediction model, to calculate a systolic pressure value and a diastolic pressure value.

Image based blood pressure monitoring method

An image coding method including constructing a plurality of edge models with a Forward Discrete Cosine Transform (FDCT) algorithm; creating adjustment equations each matching one of the edge models; capturing an image comprising pixels; selecting the pixels of the image to define image blocks; detecting by block-edge detection (BED) a pattern collectively exhibited by the pixels in the each of the image blocks and then comparing the detected pattern with patterns of the edge models; changing the patterns of the image blocks to the patterns of the edge models and adjusting the dominating coefficient by the adjustment factor after determining that the patterns of the image blocks approximate to the patterns of the edge models; and performing a coding process on the edge models by LLEC to generate a compressed image corresponding to the edge models. An embedded system is applicable to the image coding method.

Embedded system using image coding method

A license plate recognition system and method recognizes a license plate attached to a vehicle and bears a license plate number. The system includes a database module, a reducing module, an image capturing unit, a sorting unit, and a recognizing module. The database module stores license plate position sample type-related data and license plate number sample type-related data whose quantity is reducible by the reducing module. The image capturing unit captures license plate images at different imaging resolutions to generate license plate position image eigenvalue and license plate number image eigenvalue. The sorting unit computes the license plate position image eigenvalue to generate license plate position type-related data and computes the license plate number image eigenvalue to generate license plate number type-related data. The recognizing module compares the license plate number type-related data with the license plate number sample type-related data, thereby recognizing the license plate number.

License plate recognition system and method

The present invention discloses a vehicle tracking system and method, and the tracking method comprises the steps of capturing a bright object from an image by the bright object segmentation; labeling the bright object by a connected component labeling method and forming a connected component object; identifying, analyzing and combining the characteristics of the connected component object to form a lamp object by the bright object recognition; tracking the trajectory of the lamp object by a multi-vehicle tracking method; and identifying the type of a vehicle having the lamp object by the vehicle detection/recognition and counting the number of various vehicles.

Vehicle tracking system and tracking method thereof

The invention relates to a method for image processing, which can be used to calibrate the background quickly. When the external environment is changed due to the switch of light, the color of background is calibrated quickly, and the background can be updated together. The method not only is used to update the background, but also can be used to eliminate the convergence of background again.

Growing community of inventors

Hsinchu, Taiwan

Bing-Fei Wu