STM Journals Peer-reviewed research from Consortium e-Learning Network

Journal of Software Engineering Tools & Technology Trends

Skin Tone based Face Detection using Nonlinear Transformation and Its DSP Implementation

  1. M V Ganeswara Rao
  2. P R Kumar
  3. A M Prasad

Abstract

AbstractThe face detection plays key role in the facial recognition algorithms, which are used in many applications such as security surveillance, person identification etc. This paper proposes a new skin tone centred face detection algorithm to improve detection accuracy and detection speed. A nonlinear transformation technique used to decrease false alarms in skin tone detection stage. The proposed algorithm has been implemented on Taxas Instruments TMS320DM642 DSP processor and our algorithm increased detection accuracy by 6.2% and detection performance by 36% compared with the state of the art face detection systems. Keywords: Face detection, skin tone, nonlinear transform, digital signal processorCite this ArticleRao MVG, Kumar PR, Prasad AM. Skin Tone based Face Detection using Nonlinear Transformation and Its DSP Implementation. Journal of Software Engineering Tools & Technology Trends. 2017; 4(3): 9–16p.
Support