This paper presents the image-quality-guided strategy for optimization of bicubic interpolation and interpolated scan conversion algorithms. This strategy uses feature selection through line chart data visualization technique and first index of the minimum absolute difference between computed scores and ideal scores to determine the image quality guided coefficient k that changes all sixteen BIC coefficients to new coefficients on which the OBIC interpolation algorithm is based. Perceptual evaluations of cropped sectored images from Matlab software implementation of interpolated scan conversion algorithms are presented. Also, IQA metrics-based evaluation is presented and demonstrates that the overall performance of the OBIC algorithm is 92.22% when compared with BIC alone, but becomes 57.22% with all other methods mentioned.
Download