Coarse classification of on-line Chinese characters via structure feature-based method

A hybrid strategy of using (i) locally linear embedding for nonlinear dimensionality reduction of high dimensional data and (ii) support vector machines for classification of the resultant features is proposed as a robust methodology for process monitoring. Illustrative examples substantiate the methodology vis-à-vis current practice.

Download Full-text

On-line recognition of limited0vocabulary Chinese characters using multiple concolllutional neural networks

1993 IEEE International Symposium on Circuits and Systems ◽

10.1109/iscas.1993.693182 ◽

2005 ◽

Cited By ~ 4

Author(s):

Quen-Zong Wu ◽

Y. Le Cun ◽

L.D. Jackel ◽

Sor-SHenn Jeng

Keyword(s):

Neural Networks ◽

Chinese Characters ◽

On Line

Download Full-text

Classification of VHR remote sensing images using local feature-based attribute profiles

2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS) ◽

10.1109/igarss.2017.8127060 ◽

2017 ◽

Cited By ~ 4

Author(s):

Minh-Tan Pham ◽

Sebastien Lefevre ◽

Erchan Aptoula ◽

Bharath Bhushan Damodaran

Keyword(s):

Remote Sensing ◽

Local Feature ◽

Remote Sensing Images ◽

Feature Based

Download Full-text

On-Line Chinese Character Recognition Via a Representation of Spatial Relationships Between Strokes

International Journal of Pattern Recognition and Artificial Intelligence ◽

10.1142/s0218001497000159 ◽

1997 ◽

Vol 11 (03) ◽

pp. 329-357 ◽

Cited By ~ 5

Author(s):

Ju-Wei Chen ◽

Suh-Yin Lee

Keyword(s):

Character Recognition ◽

Chinese Character ◽

Spatial Relationship ◽

Spatial Relations ◽

Chinese Characters ◽

Spatial Relationships ◽

Matching Problem ◽

Structural Rules ◽

Chinese Character Recognition ◽

On Line

Chinese characters are constructed by basic strokes based on structural rules. In handwritten characters, the shapes of the strokes may vary to some extent, but the spatial relations and geometric configurations of the strokes are usually maintained. Therefore these spatial relations and configurations could be regarded as invariant features and could be used in the recognition of handwritten Chinese characters. In this paper, we investigate the structural knowledge in Chinese characters and propose the stroke spatial relationship representation (SSRR) to describe Chinese characters. An On-Line Chinese Character Recognition (OLCCR) method using the SSRR is also presented. With SSRR, each character is processed and is represented by an attribute graph. The process of character recognition is thereby transformed into a graph matching problem. After careful analysis, the basic spatial relationship between strokes can be characterized into five classes. A bitwise representation is adopted in the design of the data structure to reduce storage requirements and to speed up character matching. The strategy of hierarchical search in the preclassification improves the recognition speed. Basically, the attribute graph model is a generalized character representation that provides a useful and convenient representation for newly added characters in an OLCCR system with automatic learning capability. The significance of the structural approach of character recognition using spatial relationships is analyzed and is proved by experiments. Realistic testing is provided to show the effectiveness of the proposed method.

Download Full-text

Detection of Roller Bearing Conditions Using Mechatronics Approach

10.1115/imece1999-0732 ◽

1999 ◽

Author(s):

T. I. Liu ◽

F. Ordukhani

Keyword(s):

Neural Networks ◽

Roller Bearing ◽

Diagnostic System ◽

Feature Selection Technique ◽

Vibration Signals ◽

Selection Technique ◽

On Line ◽

Line Classification ◽

Nonlinear Pattern

Abstract An on-line monitoring and diagnostic system is needed to detect faulty bearings. In this work, by applying the feature selection technique to the data obtained from vibration signals, six indices were selected. Artificial neural networks were used for nonlinear pattern recognition. An attempt was made to distinguish between normal and defective bearings. Counterpropagation neural networks with various network sizes were trained for these tasks. The counterpropagation neural networks were able to recognize a normal from a defective bearing with the success rate between 88.3% to 100%. The best results were obtained when all the six indices were used for the on-line classification of roller bearings.

Download Full-text

Robust text watermarking based on line shifting

Journal of Science and Technology on Information security ◽

10.54654/isj.v7i1.50 ◽

2020 ◽

Vol 7 (1) ◽

pp. 3-13

Author(s):

Alexander Kozachok ◽

Sergey Kopylov

Keyword(s):

Block Diagram ◽

Data Embedding ◽

Process Data ◽

Text Data ◽

Data Format ◽

Robust Watermark ◽

Watermark Embedding ◽

Line Spacing ◽

On Line

Abstract— This article presents an approach to protection of printed text data by watermark embedding in the printing process. Data protection is based on robust watermark embedding that is invariant to text data format converting into image. The choice of a robust watermark within the confines of the presented classification of digital watermark is justified. The requirements to developed robust watermark have been formed. According to the formed requirements and existing restrictions, an approach to robust watermark embedding into text data based on a steganographic algorithm of line spacing shifting has been developed. The block diagram and the description of the developed algorithm of data embedding into text data are given. An experimental estimation of the embedding capacity and perceptual invisibility of the developed data embedding approach was carried out. An approach to extract embedded information from images containing a robust watermark has been developed. The limits of the retrieval, extraction accuracy and robustness evaluation of embedded data to various transformations have been experimentally established.Tóm tắt— Bài báo trình bày cách tiếp cận để bảo vệ dữ liệu văn bản in bằng cách nhúng vào văn bản một đoạn thủy vân trong quá trình in. Bảo vệ dữ liệu dựa trên việc sử dụng thủy vân bền vững có khả năng chống lại sự chuyển đổi định dạng dữ liệu văn bản sang dữ liệu hình ảnh. Sau quá trình phân tích các hệ thống thủy vân số hiện có, nhận thấy việc lựa chọn một mô hình thủy vân bền vững là hợp lý. Do yêu cầu thực tế và các hạn chế của phương pháp nhúng thủy vân vào dữ liệu văn bản hiện có, bài báo đưa ra phương pháp nhúng mới được phát triển dựa trên một thuật toán ẩn mã sử dụng cách thay đổi khoảng cách giữa các dòng trong văn bản. Bài báo đưa ra một sơ đồ khối và mô tả thuật toán nhúng thông tin vào dữ liệu văn bản. Các thực nghiệm về khả năng nhúng và khả năng che giấu thông tin với tri giác thông thường của dữ liệu nhúng cũng được trình bày. Bài báo cũng nêu cách tiếp cận để trích xuất thông tin được nhúng từ các hình ảnh có chứa thủy vân bền vững. Bên cạnh đó, chúng tôi cũng đưa ra các giới hạn về khả năng ứng dụng của phương pháp dựa trên các thực nghiệm, các đánh giá về độ chính xác của việc trích xuất được dữ liệu và độ mạnh của phương pháp nhúng mới này đối với các phép biến đổi ảnh khác nhau.

Download Full-text