Noise reduction of radiochromic film: Median filter processing of subtraction image

Abstract

Pre-ultraviolet rays exposure is a useful method to reduce non-uniformity error of radiochromic films. However, dust and scratch noises such as spike noise disturb precise measurement. To reduce these noises, median filter processing is applied for pre-subtraction and subtraction images. To reduce non-uniformity error of the thickness unevenness of Gafchromic EBT film, ultraviolet rays were exposed to correct data. There were three kinds of images obtained: first ultraviolet exposure image, second ultraviolet exposure image and the subtraction image of both. Median filer processing was performed on all these images. Eleven kinds of median filter radius factors (0.0 to 5.0) were applied using image analysis software. Data and graphs were then estimated. The maximum pixels value of dust was 229 on the second ultraviolet exposure image of film 3. After median filter preprocessing, the pixel value of the noises were similar to the minimum value. A 2.0-radius median filer is a useful factor for processing. Noise reduction that affected data of estimated images may be applied to measure radiation doses on a variety of radiochromic films. Ultraviolet exposure and subtraction method with median filter processing enable precise measurement and high spatial resolution dose distribution.