No icon

Dissecting and Reassembling Color Correction Algorithms for Image Stitching

Dissecting and Reassembling Color Correction Algorithms for Image Stitching

Abstract:

This paper introduces a new compositional framework for classifying color correction methods according to their two main computational units. The framework was used to dissect fifteen among the best color correction algorithms and the computational units so derived, with the addition of four new units specifically designed for this work, were then reassembled in a combinatorial way to originate about one hundred distinct color correction methods, most of which never considered before. The above color correction methods were tested on three different existing datasets, including both real and artificial color transformations, plus a novel dataset of real image pairs categorized according to the kind of color alterations induced by specific acquisition setups. Differently from previous evaluations, special emphasis was given to effectiveness in real world applications, such as image mosaicing and stitching, where robustness with respect to strong image misalignments and light scattering effects is required. Experimental evidence is provided for the first time in terms of the most recent perceptual image quality metrics, which are known to be the closest to human judgment. Comparative results show that combinations of the new computational units are the most effective for real stitching scenarios, regardless of the specific source of color alteration. On the other hand, in the case of accurate image alignment and artificial color alterations, the best performing methods either use one of the new computational units, or are made up of fresh combinations of existing units.

Existing System:

We relied mainly on the protocol described (the most complete comparison of color correction methods for stitching to date, here significantly extended), which assumes no known color transform model, this being the most general and sensible way to address the problem. Indeed, most recent datasets avoid to refer to any particular image acquisition conditions, i.e. operating setups (e.g. single vs multiple cameras, fixed vs changing camera parameters), due to the impossibility for the common user to have a fully controlled environment and the right level of knowledge and expertise. Nevertheless, we verified experimentally that the results of the best color correction methods are virtually uncorrelated with the acquisition setup and therefore with the associated color alteration.

Proposed System:

We introduced a new compositional framework for classifying color correction methods in terms of ME/PA pairs. This framework is completely general and comprehensive, and allows for a clearer analysis of color correction methods, providing a deeper insight into their properties. We revisited and categorized 15 of the existing color correction methods according to this framework, identifying and combining pairwise their MEs and PAs to design new methods never considered before.

Comment As:

Comment (0)