A Secure and High-Capacity Data-Hiding Method Using Compression, Encryption and Optimized Pixel Value Differencing

ABMC-DH: An Adaptive Bit-Plane Data Hiding Method Based on Matrix Coding

“…To demonstrate the effectiveness and superiority of the proposed ABMC-DH method, we conducted comparison experiments against methods by Crandall [13], Zhang et al [14], Cao et al [15], and Shukla et al [30] with our method from two aspects: embedding capacity (or E R ) and visual quality (PSNR). Nine grayscale test images shown in the Fig.…”

“…For more embedding capacity, Crandall's [13] and Zhang et al's [14] schemes need to provide more binary bits to participate in the hiding procedure, and there is a great impact on PSNR when the number of LSBs exceeds four. In Shukla et al's scheme [30], for each pixel the number of LSBs replaced is greater or equal to three. Therefore, though the method can provide a large embedding capacity, it is inferior to ours in term of PSNR under the same embedding rate.…”

“…Finally, through four typical grayscale test images, which include the two smooth images of ''Lena'' and ''Airplane'', as well as two complex images ''Baboon'' and ''Pepper'', we compare data hiding performance under the same embedding capacity between the proposed ABMC-DH method and four representative studies: Crandall [13], Zhang et al [14], Cao et al [15], and Shukla et al [30]. The experimental results are plotted on the coordinate plane using line charts, and the graph is shown in Fig.…”

“…8, in which the x-axis represents hiding capability rate, the y-axis represents image visual quality, and the points on the coordinate plane represent a set of experimental data. Our results are represented by a red curve with circle marks, Crandall's scheme [13] is represented by a blue curve with cross marks, Zhang et al's scheme [14] is represented by a yellow curve with triangle marks, Cao et al's [15] results are represented by an green curve with cross marks, and Shukla et al's [30] results are presented by a black curve with square marks.…”

“…A data hiding method based on block truncation coding using pixel pair matching technique was proposed by Hong [24], which classified the compressed AMBTC blocks into smooth and complex blocks using a predefined threshold, and adopted the bitmap replacement strategy in smooth blocks and Chen et al's method [25] in complex blocks, and moreover, two quantization levels were recalculated to minimize distortion and log ∂ 2 data was embedded into two quantization levels using a pixel pair matching technique, thus the method achieves better image quality with a high payload and comparable image quality with low payload compared to previous schemes. A high-capacity data hiding method was proposed by Shukla et al [30], which uses arithmetic coding to compress secret data to provide extra capacity and encrypts the compressed data by AES for security, and then divides the cover image into non-overlapping pixel blocks size of 3 × 3 or 2 × 2, using a PVD+LSB method to embed equal or more than 3-bits into each pixel. Finally, an optimization strategy achieves the smallest distortion compared to previous methods.…”