Miroslav Goljan

Lossless data embedding--new paradigm in digital watermarking

One common drawback of virtually all current data embedding methods is the fact that the original image is inevitably distorted due to data embedding itself. This distortion typically cannot be removed completely due to quantization, bit-replacement, or truncation at the grayscales 0 and 255. Although the distortion is often quite small and perceptual models are used to minimize its visibility, the distortion may not be acceptable for medical imagery (for legal reasons) or for military images inspected under nonstandard viewing conditions (after enhancement or extreme zoom). In this paper, we introduce a new paradigm for data embedding in images (lossless data embedding) that has the property that the distortion due to embedding can be completely removed from the watermarked image after the embedded data has been extracted. We present lossless embedding methods for the uncompressed formats (BMP, TIFF) and for the JPEG format. We also show how the concept of lossless data embedding can be used as a powerful tool to achieve a variety of nontrivial tasks, including lossless authentication using fragile watermarks, steganalysis of LSB embedding, and distortion-free robust watermarking.

Detecting LSB steganography in color, and gray-scale images

We describe a reliable and accurate method for detecting least significant bit (LSB) nonsequential embedding in digital images. The secret message length is derived by inspecting the lossless capacity in the LSB and shifted LSB plane. An upper bound of 0.005 bits per pixel was experimentally determined for safe LSB embedding.

Determining Image Origin and Integrity Using Sensor Noise

In this paper, we provide a unified framework for identifying the source digital camera from its images and for revealing digitally altered images using photo-response nonuniformity noise (PRNU), which is a unique stochastic fingerprint of imaging sensors. The PRNU is obtained using a maximum-likelihood estimator derived from a simplified model of the sensor output. Both digital forensics tasks are then achieved by detecting the presence of sensor PRNU in specific regions of the image under investigation. The detection is formulated as a hypothesis testing problem. The statistical distribution of the optimal test statistics is obtained using a predictor of the test statistics on small image blocks. The predictor enables more accurate and meaningful estimation of probabilities of false rejection of a correct camera and missed detection of a tampered region. We also include a benchmark implementation of this framework and detailed experimental validation. The robustness of the proposed forensic methods is tested on common image processing, such as JPEG compression, gamma correction, resizing, and denoising.

Steganalysis of JPEG Images: Breaking the F5 Algorithm

In this paper, we present a steganalytic method that can reliably detect messages (and estimate their size) hidden in JPEG images using the steganographic algorithm F5. The key element of the method is estimation of the cover-image histogram from the stego-image. This is done by decompressing the stego-image, cropping it by four pixels in both directions to remove the quantization in the frequency domain, and recompressing it using the same quality factor as the stego-image. The number of relative changes introduced by F5 is determined using the least square fit by comparing the estimated histograms of selected DCT coefficients with those of the stego-image. Experimental results indicate that relative modifications as small as 10% of the usable DCT coefficients can be reliably detected. The method is tested on a diverse set of test images that include both raw and processed images in the JPEG and BMP formats.

Robust hash functions for digital watermarking

Digital watermarks have been proposed for authentication of both video data and still images and for integrity verification of visual multimedia. In such applications, the watermark has to depend on a secret key and on the original image. It is important that the dependence on the key be sensitive, while the dependence on the image be continuous (robust). Both requirements can be satisfied using special image digest functions that return the same bit-string for a whole class of images derived from an original image using common processing operations. It is further required that two completely different images produce completely different bit-strings. We discuss methods how such robust hash functions can be built. We also show how the hash bits has another application, the robust image digest can be used as a search index for an efficient image database search.

Distortion-Free Data Embedding for Images

One common drawback of virtually all current data embedding methods is the fact that the original image is inevitably distorted by some small amount of noise due to data embedding itself. This distortion typically cannot be removed completely due to quantization, bit-replacement, or truncation at the grayscales 0 and 255. Although the distortion is often quite small, it may not be acceptable for medical imagery (for legal reasons) or for military images inspected under unusual viewing conditions (after filtering or extreme zoom). In this paper, we introduce a general approach for high-capacity data embedding that is distortion-free (or lossless) in the sense that after the embedded information is extracted from the stego-image, we can revert to the exact copy of the original image before the embedding occurred. The new method can be used as a powerful tool to achieve a variety of non-trivial tasks, including distortion-free robust watermarking, distortion-free authentication using fragile watermarks, and steganalysis. The proposed concepts are also extended to lossy image formats, such as the JPG.

Images with self-correcting capabilities

In this paper, we introduce two techniques for self-embedding an image in itself as a means for protecting the image content. After self-embedding, it is possible to recover portions of the image that have been cropped out, replaced, damaged, or otherwise tampered without accessing the original image. The first method is based on transforming small 8/spl times/8 blocks using a DCT, quantizing the coefficients, and carefully encoding them in the least significant bits of other, distant squares. This method provides very high quality of reconstruction but it is very fragile. The quality of the reconstructed image areas is roughly equivalent to a 50% quality JPEG compressed original. The second method uses a principle similar to differential encoding to embed a circular shift of the original image with decreased color depth into the original image. The quality of the reconstructed image gradually degrades with increasing amount of noise in the tampered image. The first technique can also be used as a fragile watermark for image authentication, while the second technique can be classified as a semi-robust watermark.

Steganalysis based on JPEG compatibility

In this paper, we introduce a new forensic tool that can reliably detect modifications in digital images, such as distortion due to steganography and watermarking, in images that were originally stored in the JPEG format. The JPEG compression leave unique fingerprints and serves as a fragile watermark enabling us to detect changes as small as modifying the LSB of one randomly chosen pixel. The detection of changes is based on investigating the compatibility of 8x8 blocks of pixels with JPEG compression with a given quantization matrix. The proposed steganalytic method is applicable to virtually all steganongraphic and watermarking algorithms with the exception of those that embed message bits into the quantized JPEG DCT coefficients. The method can also be used to estimate the size of the secret message and identify the pixels that carry message bits. As a consequence of our steganalysis, we strongly recommend avoiding using images that have been originally stored in the JPEG format as cover-images for spatial-domain steganography.

Invertible authentication watermark for JPEG images

We present two new invertible watermarking methods for authentication of digital images in the JPEG format. While virtually all previous authentication watermarking schemes introduced some small amount of non-invertible distortion in the image, the new methods are invertible in the sense that, if the image is deemed authentic, the distortion due to authentication can be completely removed to obtain the original image data. The first technique is based on lossless compression of biased bit-streams derived from the quantized JPEG coefficients. The second technique modifies the quantization matrix to enable lossless embedding of one bit per DCT coefficient. Both techniques are fast and can be used for general distortion-free (invertible) data embedding. The new methods provide new information assurance tools for integrity protection of sensitive imagery, such as medical images or high-importance military images viewed under non-standard conditions when usual criteria for visibility do not apply.

On estimation of secret message length in LSB steganography in spatial domain

In this paper, we present a new method for estimating the secret message length of bit-streams embedded using the Least Significant Bit embedding (LSB) at random pixel positions. We introduce the concept of a weighted stego image and then formulate the problem of determining the unknown message length as a simple optimization problem. The methodology is further refined to obtain more stable and accurate results for a wide spectrum of natural images. One of the advantages of the new method is its modular structure and a clean mathematical derivation that enables elegant estimator accuracy analysis using statistical image models.