Abstract: A signal embedder hides auxiliary data in a media signal such that the auxiliary data is humanly imperceptible yet recoverable by an automated auxiliary data reader. The embedding method comprises segmenting the media signal into regions, determining statistics for the regions, and adapting quantization bins for each region based on the statistics calculated for the region. To hide auxiliary data in the regions, the method quantizes signal characteristics in the regions into the quantization bins adapted for the regions. The quantization bins correspond to auxiliary data symbols and the signal characteristics are quantized into selected bins depending on the auxiliary data symbol to be embedded in the signal characteristics. A compatible reading method performs a similar adaptive process to define the quantization bins before mapping signal characteristics into the adapted bins to extract the hidden data.

Abstract: Certain forms of distortion make it difficult to recover hidden data embedded in an audio or image signal by quanitzation techniques. To compensate for this distortion, an embedded data reader analyzes a statistical distribution (e.g., a histogram) of feature samples in an audio or image signal suspected of having hidden auxiliary data to derive an estimate of quantizers used to encode a reference signal. The estimated quantizers then recover the reference signal, and the reader uses the reference signal to determine and compensate for geometric or temporal distortion, like spatial scaling and rotation of image data, and time scale and speed changes of audio data. After compensating for such distortion, the reader can then more accurately recover hidden message data using quantization techniques to extract the message. The reference signal is preferably repeated in blocks of the image or audio data to enable synchronization at many points in an image or audio data stream.

Abstract: A fluid seal having differing fluid levels is provided on opposite sides of a downwardly opening cup-shaped container that is longitudinally lapped by an upwardly opening annular container that is sealed to a vessel such that the lip of the cup-shaped container is submerged. The fluid seal creates a hydrostatic pressure barrier between ambient and a vessel, for example at a rotating vertical shaft. Relatively high speeds are permitted without leakage due to a number of internal horizontal and vertical baffles on one or more walls of the annular container. Barrier fluid losses due to evaporation or the like can be replaced in an embodiment wherein condensed process fluid is used as the barrier fluid, and wherein excess barrier fluid is returned to the vessel while maintaining a differential pressure barrier via the seal.

Abstract: The present invention detects the presence of a watermark in-an image by using a multi-step process. First, the image is examined to determine which regions of the image have characteristics such that there is a high probability that a watermark signal can be detected in that region of the image. Next the regions that have a high probability that a watermark can be detected (in contrast to all regions of the image) are examined to find watermark data. In order to determine the probability of finding watermark data in a particular region of an image, the amount of “variance” in the intensity of the pixels in the region is first examined. For example a region that is entirely white or entirely black has zero variance in luminance. Such a region can not carry watermark data, hence regions with zero or low variance can be eliminated from further processing.

Abstract: Steganographic calibration signals (sometimes termed “orientation signals,” “marker signals,” reference signals,” “grid signals,” etc.) are sometimes included with digital watermarking signals so that subsequent distortion of the object thereby marked (e.g., a digital image file, audio clip, document, etc.) can later be discerned and compensated-for. Digital watermark detection systems sometimes fail if the object encompasses several separately-watermarked components (e.g., a scanned magazine page with several different images, or photocopy data resulting from scanning while several documents are on the photocopier platen). Each component may include its own calibration signal, confusing the detection system. In accordance with certain embodiments, this problem is addressed by a proximity-based approach, and/or a multiple grid-based approach.

Abstract: Certain forms of distortion make it difficult to recover hidden data embedded in an audio or image signal by quanitzation techniques. To compensate for this distortion, an embedded data reader analyzes a statistical distribution (e.g., a histogram) of feature samples in an audio or image signal suspected of having hidden auxiliary data to derive an estimate of quantizers used to encode a reference signal. The estimated quantizers then recover the reference signal, and the reader uses the reference signal to determine and compensate for geometric or temporal distortion, like spatial scaling and rotation of image data, and time scale and speed changes of audio data. After compensating for such distortion, the reader can then more accurately recover hidden message data using quantization techniques to extract the message. The reference signal is preferably repeated in blocks of the image or audio data to enable synchronization at many points in an image or audio data stream.

Abstract: An embedded signal detection process determines a transformation of a media signal subsequent to the encoding of an embedded code signal into the media signal. The process performs a logarithmic sampling of the media signal to create a sampled signal in which scaling of the media signal is converted to translation in the sampled signal. It then computes the translation of the embedded code signal in the sampled signal to determine scaling of the media signal subsequent to the encoding of the embedded signal in the media signal.

Abstract: A watermark detector maps target media data into a log polar coordinate system and correlates the target media with a detection watermark to compute orientation parameters. The correlation process computes a measure of correlation for an array of potential orientation parameter candidates. Evaluating the correlation associated with these candidates, the detector selects one or more of the orientation parameters. It then proceeds to refine the correlation by using the computed orientation parameters, namely scale and rotation, to find additional parameters such as translation and differential scale.

Abstract: The present invention detects the presence of a digital watermark in an image by selecting regions within the image having a high probability of containing the watermark. The image is examined to determine which regions of the image have characteristics indicating that there is a high probability that a watermark signal can be detected in that region of the image. The regions that have a high probability that a watermark can be detected (in contrast to all regions of the image) are examined to find watermark data. Probability factors used to select detection regions include: requiring a minimum variance separation between detection blocks; requiring a minimum distance between overlapping blocks; segmenting the detection blocks into a first and second subset based on separate criteria; establishing a keep away zone to prevent selection of a detection block near an image's border; and selecting a detection block only after its neighbors met sufficient threshold requirements.

Abstract: A wavelet domain watermark encoder and decoder embed and detect auxiliary signals in a media signal, such as a still image, video or audio signal. A watermark orientation signal is embedded in a wavelet decomposed signal to facilitate detection of the watermark in a geometrically distorted version of the embedded signal.

Abstract: Synthesizing of video frames that have been dropped by a video encoder is achieved by interpolating between decoded frames at a decoder. The method consists of successive refinement stages that increase in computational complexity. Starting with a spatio-temporal median filtering approach, each stage uses information that improves the quality of the interpolated frames, such as bit stream motion information, decoder-based motion estimation and motion-based state segmentation of regions. By using more computational resources, each of these stages results in an improved quality of interpolated video. The motion compensation techniques are based on block-based motion estimation of the kind used by block-transform based video encoders. More accurate motion estimates are obtained by using a combination of forward and backward block motion estimation.

Abstract: A hydrofoil impeller includes three impeller blades attached to a hub. Each impeller blade includes a flat section and a curved section which terminates in a tip portion. The flat section has a root portion proximate the hub and which forms a predetermined root angle with respect to a plane perpendicular to the vertical axis of the impeller. The curved section is separated from the flat portion by a bend line. The curved section is bent downwardly from the plane of the flat section along the bend line. The bend line is oriented with respect to the impeller blade such that the blade tip angle is changed as the bend is formed. Accordingly, the bend is used to set the proper tip angle. The blade angle increases from the tip inward to the bend line, then becomes a constant where attached to the hub. The camber increase from the tip inward to the bend line, then falls to zero where attached to the hub.