Wistron Corporation

Ming-Che Ho

Ching-An Cho

You-Jyun Syu

Chong-Wei Li

Yu-Wen Huang

Yi-Sheng Chao

Shih-Hao Kuo

Yueh-Chi Hung

Shang-Min Yeh

A method, an image processing device, and a system for generating a disparity map are proposed. The method includes the following steps. Images of a specific scene are captured by two image sensors to generate two original images. A shrinking process is performed on the two original images to generate two shrunk images. A shrunk disparity map is generated by using the two shrunk images. A magnifying process is performed on the shrunk disparity map to generate a magnified disparity map. Whether each magnified disparity value of the magnified disparity map is reliable is determined so as to accordingly generate a refined disparity map.

Method, image processing device, and system for generating disparity map

A target tracking method and system adaptable to multi-target tracking include performing global-search detection on a current image to obtain candidates of the current image. Association between the candidates and a tracked target is performed to determine similarity between the candidates and the tracked target and to give corresponding similarity values to the candidates. The candidate with a maximum similarity value is defined as an associated candidate of the tracked target. Candidates with non-zero similarity values and primary-object classification other than the associated candidate are filtered out and defined as filtered candidates. New tracked targets are generated according to the associated candidate and the filtered candidates.

Target tracking method and system adaptable to multi-target tracking

A method, a system, and a computer-readable recording medium for long-distance person identification are provided. The method is applicable to a system having an image capturing device and a depth sensor and includes the following steps. An image of a user is captured by using the image capturing device to generate a user image, and depth information of a user is detected by using a depth sensor to generate user depth information. Soft biometric features of the user are obtained according to the user image and the user depth information, where the soft biometric features include silhouette information and human body features. A soft biometric feature similarity of the user is calculated based on the soft biometric features by using registered information of registered users so as to output a person identification result accordingly.

Method, system, and computer-readable recording medium for long-distance person identification

An analysis method of lane stripe images, an image analysis device and a non-transitory computer readable medium thereof are provided to perform steps of: setting a reference point as a center to recognize the lane stripe image in a plurality of default directions; defining a plurality of preset sections onto the lane stripe image and determining a characteristic value of the lane stripe image in each of the preset sections whenever the lane stripe image is recognized in one of the default directions; determining a first feature parameter according to the characteristic values of the lane stripe image in the preset sections when the lane stripe image is recognized in at least one of the default directions; and determining an actual lane parameter of the lane stripe image according to at least the first feature parameter.

Growing community of inventors

Taipei, Taiwan

Ming-Che Ho