A Calibration Approach for Accuracy Infrared Temperature

Abstract

The measurement of temperature of infrared image is widely used in the test field which bring much convenience. In the measuring temperature system, measurement accuracy is an imperative problem. The accuracy is influenced by many factors. The calibration based on distance segmentation parameters will be solved and the effect of other factors will be compared in this paper. Grayscale is obtained by infrared Camera. The grayscale will be transformed to temperature data. The data is calibrated by specific method. In this paper, two methods will be compared. The first one is polynomial fitting. Second one is neural network fitting. The accuracy of two methods are discussed. The experiment show that neural network fitting has much advantages and more accurate than the traditional polynomial fitting.