International Refereed Journal of Engineering and Science (IRJES) ISSN (Online) 2319-183X, (Print) 2319-1821 Volume 6, Issue 3 (March 2017), PP.26-30 www.irjes.com 26 | Page Conceptual Framework for Anthropomorphic Simulation of Human Face for Interactive Therapeutic Telepresence Applications Georgi Petrov 1 , Polina Mihova 1 , Filip Andonov 1 New Bulgarian University, Montevideo 21 Str, ABSTRACT. This publication is dedicated to the experimental development of conceptual framework for interactive direct marketing network based on simulated anthropomorphic agents (human faces). Aims: to discuss the main functional open system architecture and in particular openness to track and identify the mood of the user by analyzing the captured images of his face and other features. The article presents practical results, problems and possible solutions on the particular stage of development of the prototype, including project for further investigation the recognition of standardized facial expressions of emotion (anger, fear, disgust, happiness, sadness, surprise) at a perceptual level for children on the autism spectrum. Keywords: simulated agent, prototype of human face, emotions and autism spectrum disorders (ASD). I. INTRODUCTION This publication discusses an open conceptual model development of distributed video monitoring system for recording and processing images of people / children through implementation of persons` anthropomorphic stylized and animated images for specific therapeutic applications [1,2]. One of the purposes of the system is to stimulate children with disabilities and special needs for perception of facial expressions, distinguishing emotions and etc. The chosen model should allow limited scalability of hardware and to be based on an open software and hardware solutions. The current project uses hardware platform based on minicomputer module Raspberry Pi and it`s low-budget implementations Banana Pi. The operating systems of the hardware modules allow construction of network related distributed computing architecture, based on TCP/IP and multicasting applications, where synchronization functions are based on standard network protocols. The network hub is a budgetary solution by the company MikroTik, - the only one that offers integrated options for VPN, WLAN and multicasting administrative access in a single console at a low cost. The system should support network module additions that can process and analyze camera`s and microphone`s multicasting [3] incoming information and thus allow the hardware kit upgrade in order to develop 3D modeling with temporal recording and analysis movement of children/patients complex movements. [4] 1.1. Main components of the system The basic system`s model is realized with ARM microcomputer, Linux-based solutions with open hardware and software. Author’s choice for programming language support is Python, which allows system`s performance on standard PC based devices, whish greatly expands the application areas and the future developments. The architecture choice are Raspberry Pi based solutions, but depending on the needs, objectives and available financial resources could be introduced a high-budgetarchitecture. The basic system should have at least one 1080p@30fps resolution video camera, connected to a Raspberry Pi compatible system and a host computer. The hosting software should have options to work with standard USB cameras, but the applied Pi module significantly expands the architecture potential in terms of building multi-element video monitoring systems [5], consisting of two or more video cameras with synchronous separate video streams transmitting such as multicast messages through the local network. The primary connecting device`s protocol is an IP-based solution with specific recommendation for a wired LAN connection between the individual Pi-based cameras and the network hub. In our project, we used the MikroTik company`s product RB951G-2HnD. This type of device has number of features that successfully replace Cisco solutions whose cost exceeds ten prototypes without sacrificing the performance. The network hub can be of any type but should support a VPN, WLAN and Multicasting. Each camera is connected to the developed LAN for continuously video and audio stream broadcasting in multicasting message`s mode [6]. The selected architecture allows associating cameras and simultaneous traffic interception. The network latency is negligible due to the fact that the LAN is used locally.