Super-resolution reconstruction with photometric change of color image sequence
Jia Su-jua · Chinese Journal of Liquid Crystals and Displays · 2014
In order to realize the super-resolution reconstruction of color sequence images when the external illumination changes or the camera adjusts integration time automatically,an algorithm has been proposed.Firstly,the images' histograms of oriented gradients(HOG)are extracted as the matching characters and the correlation coefficient is used as the similarity measurements for geometric registration.Secondly,aphotometric model is established to achieve photometric registration.Finally,the algorithm of Maximum a Posteriori(MAP)is adopted to get high-resolution images.In order to show the advantage of the algorithm,the results of the average gradient,information entropy and image contrast metric(ICM)of reconstructed images are given.The results show that the average gradient and information entropy of these high-resolution images are enhanced significantly and images have more information.The algorithm is robust to photometric change and it can obviously reduce the photometric difference between the reconstructed images and improve the contrast of color images.