A Framework for Face Recognition Based on Fuzzy Minimal Structure Oscillation and Independent Component Analysis Sharmistha Bhattacharya (Halder) and Srijita Barman Roy Abstract This paper aims to provide a better accuracy for face recognition pro- cedure. This new algorithm is based on accurate feature extraction and proper classification. In this paper feature coordinate-based ICA is used for feature extraction. Pixel values of invariable coordinates (containing decisive data) for every training set are considered for analyzing through ICA. After feature extrac- tion, these values are used for fuzzy minimal structure oscillation-based classifi- cation. Proposed face recognition procedure accentuates improved classification considering the feature vectors, which is the outcome of independent component analysis of the face image. Keywords Fuzzy minimal structure oscillation ⋅ Feature coordinate ⋅ Feature coordinate-based ICA ⋅ Face recognition 1 Introduction In general topological space, the concept of minimal or m * x structure was presented by Popa and Noiri, in 2000 [1]. The vital concept of - V set and - Λ set was first introduced by Maki [2] in 1986. But in Fuzzy environment Alimohammady and M. Roohi first introduced the concept of fuzzy minimal structure [3]. A fuzzy set oscillates between at least two fuzzy open set and two fuzzy closed set. It was first introduced by Mukherjee and Halder in 2007 [4]. Fuzzy m * x Oscillation was first presented by Bhattacharya (Halder) and Roy in 2010 [5]. Image classification S. Bhattacharya (Halder) ( ✉ ) ⋅ S.B. Roy Department of Mathematics, Tripura University, Agartala, India e-mail: halder_731@rediffmail.com S.B. Roy e-mail: Srijita.barmanroy@gmail.com © Springer Nature Singapore Pte Ltd. 2018 S. Bhattacharyya et al. (eds.), Industry Interactive Innovations in Science, Engineering and Technology, Lecture Notes in Networks and Systems 11, DOI 10.1007/978-981-10-3953-9_20 209