Evaluating sensor linearity of chosen infrared sensors

Abstract. The paper describes a series of experiments conducted as part of the IRAMSWater Project, the aim of which is to establish methodologies for detecting and identifying pollutants in water bodies using aerial imagery data. The main idea is based on the hypothesis, that it is possible to identify certain types of physical, biological and chemical pollutants based on their spectral reflectance characteristics. The knowledge of these spectral curves is then used to determine very narrow spectral bands in which greatest reflectance variations occur between these pollutants. A frame camera is then equipped with a band pass filter, which allows only the selected bandwidth to be registered. In order to obtain reliable reflectance data straight from the images, the team at the Military University of Technology had developed a methodology for determining the necessary acquisition parameters for the sensor (integration time and f-stop depending on the distance from the scene and it's illumination). This methodology however is based on the assumption, that the imaging sensors have a linear response. This paper shows the results of experiments used to evaluate this linearity.