Invertible Color-to-Grayscale Conversion Using Lossy Compression and High-Capacity Data Hiding

Invertible color-to-grayscale conversion is a research issue in grayscale image colorization, which is a complex problem resulting from information loss. This paper presents an innovative invertible color-to-grayscale conversion idea by independently compressing the chromaticity plane and hiding it into the corresponding luminance plane as a watermark. Since the chromaticity and luminance planes are orthogonal in the proposed method, they can be processed efficiently without mutual influence. By using an efficient lossy compression operation, we can save more chromatic information. By using two high-capacity data hiding techniques (reversible watermarking (RW) and least significant bit (LSB) substitution), we can embed the compressed chromatic information into the luminance plane perfectly. For the purpose of integrality authentication, the luminance plane is hashed as part of the embedded information before the embedding. Experimental results have shown that higher quality of reconstructed color images can be achieved by using RW, but the quality of the synthesized grayscale drops sharply. By using LSB substitution, we can obtain high-quality synthesized grayscale and reconstructed color images simultaneously. Furthermore, we have compared the proposed LSB-based scheme with several recently reported state-of-the-art methods to validate the superiority of the proposed approach.