Investigation of Superresolution using Phase based Image Matching with Function Fitting
Budi Setiyono, Mochamad Hariadi, Mauridhi Hery Purnomo · 2012
Higher resolution image provide more detail information, so that it obtain more accurate image analysis. Many areas require high resolution image, such as medical, sensing satellite, image of the telescope and pattern recognition. This research make a process to obtain high resolution images, known as superresolution. This superresolution using a series of images in the same scene as the reference image. Two main stages in the super resolution are the registration and reconstruction. This research propose a composite between Phase-Based Image Matching (PBIM) registration, and reconstruction using structure - adaptive normalized convolution algorithm (SANC) and projection onto convext sets algorithm (POCs). PBIM was used to estimate translational registration stage. We used the function fitting around the peak point, to obtain sub pixel accurate shift. The results of this registration were used for reconstruction. Two registration method and reconstruction algorithms have been tested to obtain the most appropriate composite by measuring the value of peak signal to noise ratio (PSNR). In determining the effect of registration and reconstruction of objects which have different characteristics, we used some images that contain lots of texture and some other with less texture. The result showed that the composite of PBIM and reconstruction with POCs algorithm has the highest average of PSNR for both characteristics images. Images with lots of texture have PSNR average of 32.1606, while PSNR average of images with less texture was 29.99313. For every collaborative algorithm that has been tested, images with less texture have lower average of PSNR than ones with lots of texture. In this experiment, PBIM registration with function fitting has an average PSNR value of 2.88% higher than the Keren registration.