Software Design For Infrared Thermal Imaging System Based On Windows CE

With the development of the national economy, ever-increasing demands on thetemperature measurement accuracy and image quality of the infrared thermal imaginginsrument are required by the monitoring, security, medical and other applications. Inorder to improve the the infrared thermal imaging system’s imaging quality,temperature measurement accuracy, portability and the scalability, an infrared imagingsoftware system is designed based on the infrared temperature measurement theory andthermal imaging theory. Windows CE6.0operating system is adopted as the softwareplatform, because of its high real-time response capabilities, friendly man-machineinterface, easily portability between different applications. The main contents of thispaper are as follows:For the temperature measurement: the relationship between the radiationtemperature of the target object and its corresponding gray value of the infrared image isanalysied based on the theory of infrared radiation. The object’s temperature ismeasured based on the temperature calibration curve. However, with the change of thesystem’s working temperature, the temperature-gray curve of the imaging system willnot fit with the calibration curve, which greately reduces the the accuracy of thetemperature measurement. So it’s necessary to correct the infrared image system’soutput response. An output response correction method is proposed, which utilizes aflexible shutter and a temperature sensor recording the working temperature to correctthe output response through gathering the shtter’s infrared image and the workingtemperature. The experimental results indicate that this method can effectivelycompensate for the output response drift, and improve the temperature measurementaccuracy.For the infrared image processing: the characteristics of the infrared images, andthe factors that affect the image quality of the imaging system are analysed in detail.Various infrared image preprocessing algorithms including non-uniformity correction,blind pixels compensation, non-uniformity drift correction and image enhancementalgorithm are researched. As to non-uniformity correction, two-point non-uniformity correction method is adopted. As to blind pixels compensation, nearby field pixelscompensation method is utilized by the system. The non-uniformity of the infraredimging system will drift after working for a peorid of time. A one point non-uniformitycorrection method is proposed, which can effectively compensate the drift of thenon-uniformity. The image enhancement technologies based on linear transformalgorithm and histogram equalization algorithm are researched in detail. An adaptivemulti-segment linear gray transform algorithm is proposed based on the local maximumvalue of histogram of infrared image, which can dramatically enhance the contrast ofthe infrared image and suppress noise.Finally, the infrared imging system’s hardware driver and application program aredesigned and implemented. The application program is divided into base layer andapplication layer, which effectively enhance the scalability and portability of thesoftware systems. The infrared imging software system has been completed. For the320×240array infrared imaging instrument, the MRTD is no more than85mK at theenvironment temperature25℃, the display rate reaches up to25frames/sec. In therange of5℃to120℃, the temperature accuracy of the infrared imaging instrumentreaches up to±2℃.