Thermopile Array Sensor for Children Detection in Automobile Application

Abstract

Identifying children in thermal image using thermopile arrays sensor is vital for several computer vision programs, such as MATLAB, especially for detecting children left unattended in a parked automobile under extreme high temperature. Several fatal cases have been reported due to parents’ negligence, and majority has shown regretting of the mishap. The purpose of this report is to propose a dynamic background removal algorithm for human movement detection. Besides, this project aims to develop the proposed algorithm into image processing system. This can be achieved by design and setup a thermal imaging camera using thermopile array sensor, powered by microcontroller which is Raspberry Pi and then processing the thermal image by applying Background Subtraction (BS) technique to extract the foreground or human object from the background image. The expected result from this project is the extraction of foreground image (human) from background image for human detection. By adding other method such as masking method, the final outcome of the exact human detection from the thermal image can be obtained.