![]() ![]() Different from models of physics, many methods in the literature adjust the image colors based on principles of the human visual system (HVS). More sophisticated models along this direction consider the Jerlov’s water types to improve the stability.Įssentially, we are finding a color distribution on the image domain that is favorable for a human observer. These methods are sensitive to the identified veiling lights such that small perturbations on the estimated RGB values of the background trigger visually significant results. Hence, for any combination of a veiling light and a transmission map, the color-corrected image is uniquely determined. The veiling light is approximated by various types of dark-channel priors, and the estimation of the transmission map is converted to depth computation based on the Beer-Lambert law. It expresses the image colors as a convex combination of the unattenuated objects’ colors and the veiling light via a scalar transmission map. ![]() The complex factors determining the irradiance on an imaging sensor are often simplified by the Koschmieder model. ![]() We can understand this task as reversing the process of image formation. In this paper, all the underwater images are from the benchmark data set. The resulted image has enhanced contrast, balanced colors, and many image contents, e.g., the patterns on the swimming shorts, are more recognizable. In the middle, several zoomed-in regions are displayed for comparison. Figure 1: (Top) Underwater image with heavy green cast. An effective color correction method is needed to recover and enhance these details. Typically, underwater images have insufficient contrast and unbalanced color distribution, hence many image contents, such as patterns and textures are hardly recognizable for human observers. Depending on the dissolved or suspended substances, a liquid medium modifies the spectral power distribution of the transmitted light, such that a strong bluish or greenish color cast dominates the acquired underwater image, e.g., Figure 1 (Top). ![]() Water absorbs light similarly to an optical filter but with higher variations and complexities. Based on image quality metrics designed for underwater conditions, we compare with some state-of-art approaches to show that the proposed method has consistently superior performances. We analyze and validate the proposed model by various numerical experiments. To improve the uniformity of CIELAB, we include an approximate hue-linearization as the pre-processing and an inverse transform of the Helmholtz-Kohlrausch effect as the post-processing. The magnitude of the enhancement is hue-selective and image-based, thus our method is robust for different underwater imaging environments. For visualization purposes, we enhance the image contrast by properly rescaling both lightness and chroma without trespassing the CIELAB gamut. Understood as a long-term adaptive process, our method effectively removes the underwater color cast and yields a balanced color distribution. It presents a new variational interpretation of the complementary adaptation theory in psychophysics, which establishes the connection between colorimetric notions and color constancy of the human visual system (HVS). In this paper, we propose a novel approach for underwater image color correction based on a Tikhonov type optimization model in the CIELAB color space. ![]()
0 Comments
Leave a Reply. |