Title: Design of computer image automatic processing system based on artificial intelligence algorithm

Authors: Guoqiang You

Addresses: College of Engineering and Technology, Xi'an FanYi University, Xi'an, 710105, Shaanxi, China

Abstract: Automatic image processing systems are applied for recognising human faces in crowds, person identification, and face matching applications. The varying textures, input representation, and position impact detection accuracy and recognition. Therefore, this article introduces an automatic image processing method (AIPM) for face recognition (FR) using deep learning (DL) paradigm. This method extracts the textural features based on the image position and classifies them based on pixel mapping. Semantic (even) and uneven pixel variations are observed in the classification process. The semantic classified pixels are used for correlating different image segments that are further used for training the learning network. The uneven pixels classified using DL is discarded to prevent recognition errors. The DL paradigm verifies the pixel position and coordinate mapping between different inputs. The detection is improved based on the classified output for semantic and uneven pixels. The training is based on semantic and mapping pixels, for which the training is improvised using erroneous pixels. Therefore, precision is improved with controlled analysis complexity.

Keywords: artificial intelligence; automatic processing; face recognition; feature extraction; automatic image processing method; AIPM.

DOI: 10.1504/IJMTM.2024.139501

International Journal of Manufacturing Technology and Management, 2024 Vol.38 No.4/5, pp.321 - 341

Received: 17 Jun 2022
Accepted: 05 Sep 2022
Published online: 03 Jul 2024 *

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