Peking University

Boe Technology Group Co., Ltd.

Xerox Corporation

Yadong Mu

Bingfeng Zhou

Zhigang Fan

Francis Kapo Tse

Shuai Wang

Xiangye Wei

Qiong Wu

Jie Feng

Yiming Bai

Ge Ou

Xiaojun Tang

Tao He

Yigeng Fang

Guiyu Tian

He Tao

Nan Zhuang

Guoqiang Gong

The present disclosure provides an image processing method, apparatus, device, and computer-readable storage medium. The method includes: obtaining an image dataset, the image dataset including an image and an accompanying text related to an unseen class in the image; and generating a probability and/or distribution of the unseen class using an unseen class obtaining model, the probability and/or distribution of the unseen class including a probability that each pixel in the image is from the unseen class, a probability that the unseen class is present in the image, and a regional probability after the image is subdivided into a plurality of regions.

Image processing method, apparatus, and computer product for image segmentation using unseen class obtaining model

An electronic apparatus and a method for displaying an image on a display device are disclosed. The electronic apparatus includes a display device; an image acquisition device configured to acquire a surrounding image of the display device; and a processor configured to: determine a background image of the display device according to the surrounding image; acquire a target range and a target object in the background image; determine a target image according to the background image, the target range and the target object; and control the display device to display the target image, wherein the target image excludes the target object.

Electronic apparatus, and method for displaying image on display device

Provided are an image processing method, an image matching method, a device, and a storage medium. The image processing method includes: obtaining an image feature of an input image; determining a plurality of local image features of the image feature; determining a plurality of local feature vectors corresponding to the plurality of local image features respectively; determining a hash code of the input image based on the plurality of local feature vectors.

Image processing method with ash code on local feature vectors, image processing device and storage medium

A system and method of connectivity-based image processing to identify and extract objects in image data is described. Variations on the method may include iterative local smoothing operations and various algorithmic solutions to improve real-time processing. Variations may also include object extraction processes based on user-provided information about an object in an image.

Connectivity similarity based graph learning for interactive multi-label image segmentation

What is disclosed is a novel system and method for text vectorization for bitmap images with reduced artificial discontinuities. Dominant points are identified in a bitmap character image. An initial curve is fitted to edge points of the character image in a vicinity of a selected dominant point. A set of boundary parameters in a vicinity of the selected dominant point are estimated based upon the initial curve. The selected dominant point is then classified as one of a sharp dominant point and a smooth dominant point based upon the boundary parameters or alternatively upon predefined classifications produced by an optical character recognition process. Curves are fitted between the selected dominant point and adjacent dominant points. The fitted curves maintain the estimated boundary parameters in the vicinity of smooth dominant points. A vectorized representation of the text character image based upon the fitted curves is produced as output.

Between-segment discontinuity reduction for text vectorization using dominant point classification

Systems and methods are described that facilitate dominant point detection for text in a scanned document. The dominant points are classified as 'major' (e.g., structural) and 'minor' (e.g., serif). A set of rules or parameters for each character is determined off-line. During the text vectorization, OCR is performed and the rules (parameters) associated with the recognized character are selected. Both major and minor dominant points are detected as a maximization process with the parameter set. For minor dominant points, additional processes are optionally employed.

Text vectorization using OCR and stroke structure modeling

Growing community of inventors

Beijing, China

Yadong Mu