Title: Application study on image recognition algorithm in visual service of deaf mute under complex environment

Authors: Xiaoyi Yang; Qian Wu; Xinmei Deng

Addresses: School of Education, Chongqing Normal University, Chongqing, 401331, China ' School of Computer Science and Engineering, Chongqing University of Technology, Chongqing, 400054, China ' Library, Chongqing University of Science and Technology, Chongqing, 401331, China

Abstract: In order to benefit the disadvantaged deaf people and eliminate the communication disorder between the hearing and deaf mute to get harmony and symbiosis really, the visual identification and service research of deaf people has been paid more and more attention by experts and scholars. The paper presented an improved advanced local binary pattern (ALBP) algorithm for face feature recognition. Aimed at being not ideal in recognition effect resulted in ignoring the contrast between the local image in traditional face feature recognition algorithm and discarding some important texture feature information, it first made the illumination normalisation pre-processing of face images under different illumination conditions to control the illumination changes in a certain range, then the improved local binary pattern (LBP) algorithm is used to map the contrast value of the local area pixels into an interval value, so that it made the image own the illumination invariance, thereupon then it could better identify the face features. The experimental results show that the presented ALBP algorithm can be better applied to the visual service system of the deaf.

Keywords: disadvantaged people; visual service system; deaf people; communication disorder; ALBP algorithm; face recognition; feature recognition algorithm; texture feature; feature information; image process; illumination normalisation; illumination invariance.

DOI: 10.1504/IJSTM.2019.101903

International Journal of Services Technology and Management, 2019 Vol.25 No.5/6, pp.530 - 541

Received: 19 Jul 2017
Accepted: 17 Oct 2017
Published online: 30 Aug 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article