Since the refractive index of materials commonly used for lens depends on the wavelengths of light, practical camera optics fail to converge light to a single point on an image plane. Known as chromatic aberration, this phenomenon distorts image details by introducing magnification error (lateral chromatic aberration), defocus blur (axial chromatic aberration), and color fringes. Though achromatic and apochromatic lens designs reduce chromatic aberration to a degree, they are complex and expensive and they do not offer a perfect correction. In this work, we propose a new post-capture processing scheme designed to overcome these problems computationally. Specifically, the proposed solution is comprised of “chromatic aberration-tolerant” demosaicking algorithm and post-demosaicking chromatic aberration correction.
Chromatic Aberration-Tolerant Demosaicking
One challenge of designing a demosaicking method for cameras likely to be subject to chromatic aberration is the fact that cross-color correlation is significantly weakened. As a result, most demosaicking methods fail in presence of chromatic aberration—output images suffer from severe zippering artifacts. The key challenge for designing “chromatic aberration-tolerant demosaicking” is to recover the color image despite relaxing the correlation between color components. We developed a Posterior Sparsity-Directed Demosaicking (PSDD) method that does not rely on the color correlation, instead drawing on the notion of sparse representation to overcome potential hazards of aliasing. PSDD reliably recovers the chromatic aberrated image that is formed at the image plane/detector without suffering from zippering artifacts.
Chromatic Aberration Correction
Although PSDD successfully recovers complete sensor tristimulus value, this image still suffers from lateral and axial chromatic aberration. We developed a strategy to correct the color fringes caused by the magnification error and defocus blur by recovering the highpass and the lowpass components of the images separately. Compared to the conventional approaches,
- Our method handles axial chromatic aberration. (Remagnification does not correct for defocus blur)
- Our highpass recovery method does not suffer from interpolation artifacts.
- Our method corrects color fringes even when the objects in the scene are colorful. (Most methods only effective on neutral backgrounds, where correction works only by color desaturation.)
- [Reference Code] Complete implementation of camera processing pipeline for chromatically aberrated images.
- [Reference Code] Standalone implementation of PSDD.
|(2014): Camera Processing With Chromatic Aberration. In: IEEE Transactions on Image Processing, 23 (10), pp. 4539-4552, 2014.|