|Publication number||USRE42255 E1|
|Application number||US 11/496,086|
|Publication date||Mar 29, 2011|
|Filing date||Jul 27, 2006|
|Priority date||May 10, 2001|
|Also published as||US6768815, US20020168101|
|Publication number||11496086, 496086, US RE42255 E1, US RE42255E1, US-E1-RE42255, USRE42255 E1, USRE42255E1|
|Inventors||Roger L. Woodall|
|Original Assignee||Woodall Roger L|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (44), Non-Patent Citations (3), Classifications (9), Legal Events (4)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The invention described herein may be manufactured by or for the Government of the United States of America for Governmental purposes without the payment of any royalties thereon or therefore.
This patent application is co-pending with related patent applications entitled NEURAL DIRECTORS (U.S. patent application Ser. No. 09/436,957), NEURAL SENSORS (U.S. patent application Ser. No. 09/436,956), STATIC MEMORY PROCESSOR (U.S. patent application Ser. No. 09/477,638), DYNAMIC MEMORY PROCESSOR (U.S. patent application Ser. No. 09/477,653), MULTIMODE INVARIANT PROCESSOR (U.S. patent application Ser. No. 09/641,395) and A SPATIAL IMAGE PROCESSOR (Ser. No. 09/853,932), by the same inventor as this patent application.
(1) Field of the Invention
The invention relates generally to the field of color sensors and more particularly to color sensors having neural networks with a plurality of hidden layers, or multi-layer neural networks, and further to a new neural network processor for sensing color in optical image data.
(2) Description of the Prior Art
Electronic neural networks have been developed to rapidly identify patterns in certain types of input data, or accurately to classify the input patterns into one of a plurality of predetermined classifications. For example, neural networks have been developed which can recognize and identify patterns, such as the identification of hand-written alphanumeric characters, in response to input data constituting the pattern of on and off picture elements, or “pixels”, representing the images of the characters to be identified. In such a neural network, the pixel pattern is represented by, for example, electrical signals coupled to a plurality of input terminals, which, in turn, are connected to a number of processing nodes, each of which is associated with one of the alphanumeric characters which the neural network can identify. The input signals from the input terminals are coupled to the processing nodes through certain weighting functions, and each processing node generates an output signal which represents a value that is a non-linear function of the pattern of weighted input signals applied thereto. Based on the values of the weighted pattern of input signals from the input terminals, if the input signals represent a character that can be identified by the neural network, the one of the processing nodes associated with that character will generate a positive output signal, and the others will not. On the other hand, if the input signals do not represent a character that can be identified by the neural network, none of the processing nodes will generate a positive output signal. Neural networks have been developed which can perform similar pattern recognition in a number of diverse areas.
The particular patterns that the neural network can identify depend on the weighting functions and the particular connections of the input terminals to the processing nodes. The weighting functions in, for example, the above-described character recognition neural network, essentially will represent the pixel patterns that define each particular character. Typically, each processing node will perform a summation operation in connection with values representing the weighted input signals provided thereto, to generate a sum that represents the likelihood that the character to be identified is the character associated with that processing node. The processing node then applies the non-linear function to that sum to generate a positive output signal if the sum is, for example, above a predetermined threshold value. Conventional non-linear functions which processing nodes may use in connection with the sum of weighted input signals is generally a step function, a threshold function, or a sigmoid, in all cases the output signal from the processing node will approach the same positive output signal asymptotically.
Before a neural network can be useful, the weighting functions for each of the respective input signals must be established. In some cases, the weighting functions can be established a priori. Normally, however, a neural network goes through a training phase, in which input signals representing a number of training patterns for the types of items to be classified, for example, the pixel patterns of the various hand-written characters in the character-recognition example, are applied to the input terminals, and the output signals from the processing nodes are tested. Based on the pattern of output signals from the processing nodes for each training example, the weighting functions are adjusted over a number of trials. After the neural network has been trained, during an operational phase it can generally accurately recognize patterns, with the degree of success based in part on the number of training patterns applied to the neural network during the training stage, and the degree of dissimilarity between patterns to be identified. Such a neural network can also typically identify patterns that are similar, but not necessarily identical, to the training patterns.
One of the problems with conventional neural network architectures as described above is that the training methodology, generally known as the “back-propagation” method, is often extremely slow in a number of important applications. In addition, under the back-propagation method, the neural network may result in erroneous results that may require restarting of training. Even after a neural network has been through a training phase, confidence that the best training has been accomplished may sometimes be poor. If a new classification is to be added to a trained neural network, the complete neural network must be retrained. In addition, the weighting functions generated during the training phase often cannot be interpreted in ways that readily provide understanding of what they particularly represent.
Edwin H. Land's Relinex theory of color vision is based upon “three color” experiments performed before 1959. A simple “mishap” showed that three colors were not always required to see accurate color. Land used a short and long record of brightness data (black and white transparencies) to produce color perceived by human eyes and not by photographic means. He demonstrated a perception of a full range of pastel colors using two very similar in color light sources such as yellow, at 579 nm and yellow orange, at 599 nm (“Experiments in Color Vision”, Edwin H. Land, Scientific American, Vol. 200 No. May 5, 1959). Land found that in some two record experiments all colors present were not perceived. Although Land demonstrated that two records provided color perceptions, he constructed his Retinex theory upon three records such as his long, medium and short records (An Alternative Technique for the Computation of the Designator in the Retinex Theory of Color Vision”, Edwin H. Land, Proceedings of the National Academy of Sciences, Vol. 83, 1986). The invention herein is related to human color perception discovered during Land's color vision experiments as reported in 1959.
The “Trichromatic” theory in human color vision has been accepted on and off since the time of Thomas Young in 1802 (A Vision in the Brain”, S. Zeki, Blackwell Scientific Publishing, 1993). Still and video electronic camera designs are correctly based upon the trichromatic theory but the current designs are highly subjective to color error reproduction due to changes in the ambient light color temperatures and color filtrations. The device in this invention senses color using a new “bichromatic” theory, which includes a mechanism that insures color constancy over a large range of ambient color temperatures. The use of two lightness records as used by Land in 1959 is one key to this invention.
The bichromatic theory is based upon an interpretation of a biological color process that occurs in the eyes and brain of humans and in some animals. The bichromatic theory is defined as a system that functions together under the following assumptions, accepted principles and rules of procedure, for which
It is therefore an object of the invention to provide a new and improved neural network color sensor.
It is a further object to provide a neural network color sensor in which the weighting functions may be determined a priori.
Another object of the present invention is to provide a neural network color sensor, which can be trained with a single application of an input data set.
In brief summary, the color sensor generates color information defining colors of an image, comparison of colors illuminated under two or more light sources and boundaries between different colors. The color sensor includes an input section, a color processing section, a color comparison section, a color boundary processing section and a memory processing section. The input section includes an array of transducer pairs, each transducer pair defining one of a plurality of pixels of the input section. Each transducer pair comprises at least two transducers, each generating an output having a peak at a selected color, the selected color differing as between the two transducers, and each transducer having an output profile comprising a selected function of color. The color processing section includes a plurality of color pixel processors, each receiving the outputs from the two transducers comprising the transducer pair associated with a pixel. In response, the color processing section generates a color feature vector representative of the brightness of the light incident on the pixel and a color value corresponding to the ratio of outputs from the transducers comprising the transducer pair associated with the pixel. The color boundary processing section generates a plurality of color boundary feature vectors, each associated with a pixel, each representing the difference between the color value generated by the pixel color processor for the respective pixel and color values generated by the pixel color processor for pixels neighboring the respective pixel.
The color boundary sensor produces object shape feature vectors from a function of the differences in color. This color boundary sensor can sense a colored object shape in a color background where a black and white sensing retina could not detect differences in lightness between the background and the object. The color comparator processor can measure and compare the reflective color of two objects, even when each object is illuminated by two lights of different color temperatures. The memory processor section provides a process to recognize a color, a boundary of color and a comparison of colors.
A more complete understanding of the invention and many of the attendant advantages thereto will be readily appreciated as the same becomes better understood by reference to the following detailed description when considered in conjunction with the accompanying drawings wherein corresponding reference characters indicate corresponding parts throughout the several views of the drawings and wherein:
With reference to
The color processing section 12 uses the color information signals from the input section to generate, for each pixel, a local color feature vector representative of the color of the pixel. The color processing section 12 consists of a color processor array 20 and a feature fusion network array 23. The structure and operation of the color processing section 12 will be described in detail below in connection with FIG. 2. Similarly, the color boundary processing section 13 generates, for each pixel, a local color gradient feature vector that represents the gradient of the color at the pixel. The structure and operation of the color boundary processing section 13 will be described in detail below in connection with FIG. 3. The memory processor 29 is as described in STATIC MEMORY PROCESSOR, U.S. patent application Ser. No. 09/477,638. The parallel memory processors 16 and 18 are as described for the memory processor of the MULTIMODE INVARIANT PROCESSOR (U.S. patent application Ser. No. 09/641,395). The multi-mode invariant image processor, without its input sensor, is used for both parallel memory processors 16 and 18. The possible multiple outputs of the parallel memory processor 18 are the colored input object(s) classifications. The output vector array of the parallel memory processor 16 is a Positional King Of the Mountain (PKOM) array mapped to the pixels 15(m) in the retina, which becomes a map of color classifications of each pixel. It is noted that the PKOM array is a neural network array internal to the parallel memory processor 16 and the remaining neural circuits to the normal output of the MULTIMODE INVARIANT PROCESSOR are not used. The memory processor 29 is a static memory processor and provides an output classification as a degree of color comparison.
The local color feature vectors and the local color gradient feature vectors generated for all of the pixels are processed by the processing section 14 to, for example, classify the image into one of a plurality of image classes. The processing section 14 may comprise any of a plurality of processing elements for processing the vectors generated by the color processors 12, 13 and/or 19.
Each pixel color processor 20(m) includes controlled gain amplifier (CGA) circuits 30(m)(1), 30(m)(2), which receive the color amplitude signals generated by the respective transducers 15(m)(1), 15(m)(2). Each CGA circuit 30(m)(1), 30(m)(2) generates an output adjusted by a gain control factor generated by the common control 21. The gain control factor is a function of the output of the transducer for each frequency having the highest amplitude, referred to as 15(H)(1) and 15(H)(2). The CGA circuits 30(m)(1), 30(m)(2) will normalize the respective outputs in relation to the highest amplitude output for their respective frequency. This allows each transducer pair 15(m) and their respective CGA circuit 30(m) to output differing values, which represent the color at each transducer pair 15(m) as well as the “color temperature” of the light incident on the object or retina 15. The common control 21 senses all transducer outputs for each frequency and uses the highest outputs 15(H)(1), 15(H)(2) to set each CGA circuit 30(m) in the color processor 12 to the same gain as the CGA circuits 30(H)(1), 30(H)(2) from the pixel(s) 15(m) that sensed the highest light energy in retina 15. The transducers 15(H)(1), 15(H)(2), the CGA circuits 30(H)(1), 30(H)(2) and the common control 21 operate as an automatic gain controlled loop normalizing the output signal at CGA circuit 30(H)(1). Therefore, the response of each transducer 15(m)(1) is normalized at the output of each CGA circuit 30(m)(1) relative to the output of CGA circuit 30(H)(1). It is to be noted that the transducers 15(H)(1), 15(H)(2) need not be from the same pixel 15(m), as the spectral light energy of a visual scene image at two separate frequencies is generally not the same everywhere on retina 15.
The gain controlled output of each CGA circuit 30(m)(1), 30(m)(2) is provided to a number of elements, including a respective sum circuit 33(m), a difference circuit 32(m) and the common control 21. The outputs from the CGA circuits 30(m)(1), 30(m)(2) are coupled to the difference circuit, or difference generator 32(m), which generates an output vector that is representative of the difference between the amplitudes of the outputs form the CGA circuits 30(m)(1), 30(m)(2). Accordingly, it will be appreciated that the output generated by the difference generator 32(m) corresponds to the ratio of the amplitudes of the automatic controlled gain signals from the respective transducers 15(H)(1), 15(H)(2) and the respective pixel transducer 15(m) outputs.
As noted above, the outputs from the CGA circuits 30(m)(1) and 30(m)(2) are also coupled to a sum circuit 33(m). The sum circuit 33(m) generates an output that corresponds to the sum of the amplitudes of the automatic controlled gain signal from the respective transducers 15(m)(1) and 15(m)(2), and thus represents the brightness of the light incident on the pixel defined by the transducers.
The output vector from difference circuit 32(m) is coupled to the color boundary processor 13 (FIG. 1). The difference vector from difference circuit 32(m) and the brightness vector from sum circuit 33(m) are also both coupled to a neural director 35(m) that disperses these inputs into a local color feature vector. The neural director 35(m) is preferably similar to the neural directors as described in NEURAL DIRECTOR, U.S. patent application Ser. No. 09/436,957. Neural director 35(m) is preferably established to provide an output vector with an increased dimensionality, which will aid in distinguishing between similar patterns in the input vector.
The output of the neural director 35(m) is coupled to bipolar MKOM 36(m), which is described in detail in STATIC MEMORY PROCESSOR, U.S. patent application Ser. No. 09/477,638. The bipolar MKOM 36(m) generates a number of positive and/or negative outputs M(1) through M(R), generally identified by reference numeral M(r), each of which is associated with one dimension of the feature vector input thereto. Each positive component M(r) of the output vector can have a range of values from zero up to a maximum value, which corresponds to, or is proportional to, the maximum positive element value of the input vector. The positive outputs M(r) that are associated with an input vector component having successively lower positive values, are themselves successively lower in value, thus forming a positive ranking of the vector components. Outputs M(r) that are associated with input vector components having negative values are also ranked as negative vector components in a similar manner to the positive components. The rankings for the respective input feature vectors may be global, for all of the components of the input feature vector, or they may be localized among a selected number of preferably contiguous input feature vector components. The feature vector generated by the bi-polar MKOM 36(m) is coupled to the memory processing section 14.
The outputs from CGA circuits 30(m)(1) and 30(m)(2) of all of the pixel color processors 20(m) are also coupled to the common control 21. The common control 21 includes peak sensing circuits 40(1), 40(2), each of which receives the output from the correspondingly-indexed CGA circuits 30(m)(1), 30(m)(2), and each generates an output which corresponds to the one of the outputs from the correspondingly-indexed CGA circuits 30(m)(1), 30(m)(2) with the largest signal value. The outputs from the peak circuits 40(1), 40(2) are also connected to control the gain of all of the correspondingly-indexed CGA circuits 30(m)(1), 30(m)(2).
The outputs from the CGA circuits 30(m)(1) and 30(m)(2) of all of the color pixel processors 20(m) are also connected to a sum circuit 41. The sum circuit 41 generates an output, which represents the sum of the outputs from all of the CGA circuits 30(m)(1), 30(m)(2) of all of the color pixel processors 20(m). The output provided by the sum circuit 41 represents the total intensity or power of the light incident on the retina 15. An iris control circuit 42 uses the sum circuit 41 output to control the iris 17, which normalizes the intensity of the light on retina 15.
Each window difference network 50(m) receives a local window array 57(m) of difference vectors generated by the correspondingly-indexed pixel color processor 20(m). Each window difference network 50(m), in turn, generates an output vector which represents a color acceleration vector between the difference vectors provided by the correspondingly-indexed pixel color processor 20(m) and color vectors for pixels within a predetermined area around the pixel 15(m), illustrated in
In a modification to the invention 10, each pixel can be a three transducer set 15(m). Each transducer of the set 15(m) is to be matched to the response of the human retinal color cones. The three transducer set 15(m) will produce two “transducer pairs” for each pixel 15(m) and with two color processors 12 a color retina will be produced. The retina and two parallel memory processors 16 will sense color matched to the human color perception over a wide range of ambient lighting conditions.
With reference again to
The invention provides a number of advantages. In particular, the invention provides a system for receiving an image of an object and generates, for an array of pixels of the image, color and color gradient/boundary information, in the form of feature vectors, which may be processed to, for example, classify the object into one of a plurality of object classes. The system generates the color and color gradient/boundary information using only two transducers for each pixel, in accordance with a bi-chromatic color recognition scheme, with the transducers having peak responses at selected colors 1 and 2, and a known output profile as a function of color, instead of the non-color constancy process produced in accordance with the tri-chromatic color recognition scheme.
It will be appreciated that numerous modifications may be made to the system 10. For example, the memory processing section 14 may perform processing in connection with comparisons generated for two images, using output color feature vectors generated either by the same color sensor 10 at two points in time, or output comparator vectors which are generated by two color sensors (the second being denoted by 11′ and 12′) for respective pixels 15(m) for respective images. In that case, and with reference to
In addition, the peak detector circuits 40(1), 40(2) of the common control 21 may be replaced with summing circuits that generate a sum output for controlling the CGA circuits 30(m)(1), 30(m)(2).
Preferably, the iris control 42 will generally rapidly adjust the iris in response to changes in the light intensity levels incident on the retina 15, so as to maintain the light levels incident on the transducers within a predetermined operating range. In that case, the CGA circuits 30(m)(1), 30(m)(2) may have a relatively slower response to changes in the automatic gain control signals from the control circuit 21. These differences in response will allow the slower response of normalization via the CGA circuits to maintain a steady color constancy in a scene of rapid brightness changes.
The described components of invention 10 provide the necessary components for a uniquely designed photographer's exposure and color temperature meter. A calibration of the common control network 21 provides values for exposure and color temperature data. The meter may be an independent device, i.e., a hand held meter, or it may be integrated in a camera body, either electronic or film, to provide automatic exposure and color temperature corrections. The device may also be integrated into color printers or printing presses as a color ink control.
It will be apparent that variations and modifications may be made to the invention herein described and illustrated, by those skilled in the art with the attainment of some or all of the advantages of the invention. It is also understood that the color sensor described herein may be connected to the various devices described in the referenced patent applications, wherein all the devices act in concert in a manner similar to the human eye. Therefore, it is the object of the appended claims to cover all such variations and modifications as come within the true spirit and scope of the invention.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4057708||May 10, 1976||Nov 8, 1977||Motorola Inc.||Minimum miss distance vector measuring system|
|US4396903||May 29, 1981||Aug 2, 1983||Westinghouse Electric Corp.||Electro-optical system for correlating and integrating image data from frame-to-frame|
|US4963981||Jun 21, 1989||Oct 16, 1990||Hitachi, Ltd.||Image sensor device capable of electronic zooming|
|US5245672||Mar 9, 1992||Sep 14, 1993||The United States Of America As Represented By The Secretary Of Commerce||Object/anti-object neural network segmentation|
|US5263097||Jul 24, 1991||Nov 16, 1993||Texas Instruments Incorporated||Parameter normalized features for classification procedures, systems and methods|
|US5263122||Apr 22, 1991||Nov 16, 1993||Hughes Missile Systems Company||Neural network architecture|
|US5311600||Sep 29, 1992||May 10, 1994||The Board Of Trustees Of The Leland Stanford Junior University||Method of edge detection in optical images using neural network classifier|
|US5440662||Dec 11, 1992||Aug 8, 1995||At&T Corp.||Keyword/non-keyword classification in isolated word speech recognition|
|US5446828||Mar 18, 1993||Aug 29, 1995||The United States Of America As Represented By The Secretary Of The Navy||Nonlinear neural network oscillator|
|US5524065||Jul 25, 1994||Jun 4, 1996||Canon Kabushiki Kaisha||Method and apparatus for pattern recognition|
|US5613037||Dec 21, 1993||Mar 18, 1997||Lucent Technologies Inc.||Rejection of non-digit strings for connected digit speech recognition|
|US5621863||Jun 7, 1995||Apr 15, 1997||International Business Machines Corporation||Neuron circuit|
|US5629870||May 31, 1994||May 13, 1997||Siemens Energy & Automation, Inc.||Method and apparatus for predicting electric induction machine failure during operation|
|US5666467||Feb 25, 1994||Sep 9, 1997||U.S. Philips Corporation||Neural network using inhomogeneities in a medium as neurons and transmitting input signals as an unchannelled wave pattern through the medium|
|US5680481||Jun 9, 1995||Oct 21, 1997||Ricoh Corporation||Facial feature extraction method and apparatus for a neural network acoustic and visual speech recognition system|
|US5687291||Jun 27, 1996||Nov 11, 1997||The United States Of America As Represented By The Secretary Of The Army||Method and apparatus for estimating a cognitive decision made in response to a known stimulus from the corresponding single-event evoked cerebral potential|
|US5712959||Jul 7, 1995||Jan 27, 1998||Streit; Roy L.||Neural network architecture for non-Gaussian components of a mixture density function|
|US5719480||Oct 18, 1993||Feb 17, 1998||Minister Of National Defence Of Her Majesty's Canadian Government||Parametric control device|
|US5724487||Jul 7, 1995||Mar 3, 1998||Streit; Roy L.||Neural network for maximum likelihood classification with supervised and unsupervised training capability|
|US5745382||Aug 31, 1995||Apr 28, 1998||Arch Development Corporation||Neural network based system for equipment surveillance|
|US5790758||Jul 7, 1995||Aug 4, 1998||The United States Of America As Represented By The Secretary Of The Navy||Neural network architecture for gaussian components of a mixture density function|
|US5842194||Jul 28, 1995||Nov 24, 1998||Mitsubishi Denki Kabushiki Kaisha||Method of recognizing images of faces or general images using fuzzy combination of multiple resolutions|
|US5845271||Jan 26, 1996||Dec 1, 1998||Thaler; Stephen L.||Non-algorithmically implemented artificial neural networks and components thereof|
|US5850470||Aug 30, 1995||Dec 15, 1998||Siemens Corporate Research, Inc.||Neural network for locating and recognizing a deformable object|
|US5852815||May 15, 1998||Dec 22, 1998||Thaler; Stephen L.||Neural network based prototyping system and method|
|US5852816||May 15, 1998||Dec 22, 1998||Thaler; Stephen L.||Neural network based database scanning system|
|US5887087||Apr 23, 1997||Mar 23, 1999||Fuji Photo Film Co., Ltd||Image reading apparatus|
|US5974163||Dec 12, 1996||Oct 26, 1999||Nec Corporation||Fingerprint classification system|
|US6014653||May 15, 1998||Jan 11, 2000||Thaler; Stephen L.||Non-algorithmically implemented artificial neural networks and components thereof|
|US6028608||May 9, 1997||Feb 22, 2000||Jenkins; Barry||System and method of perception-based image generation and encoding|
|US6038338||Feb 3, 1997||Mar 14, 2000||The United States Of America As Represented By The Secretary Of The Navy||Hybrid neural network for pattern recognition|
|US6105015||Feb 3, 1997||Aug 15, 2000||The United States Of America As Represented By The Secretary Of The Navy||Wavelet-based hybrid neurosystem for classifying a signal or an image represented by the signal in a data system|
|US6192360||Jun 23, 1998||Feb 20, 2001||Microsoft Corporation||Methods and apparatus for classifying text and for building a text classifier|
|US6278799||Jan 24, 2000||Aug 21, 2001||Efrem H. Hoffman||Hierarchical data matrix pattern recognition system|
|US6301572||Dec 2, 1998||Oct 9, 2001||Lockheed Martin Corporation||Neural network based analysis system for vibration analysis and condition monitoring|
|US6429812||May 30, 2000||Aug 6, 2002||Steven M. Hoffberg||Mobile communication device|
|US6469804||Nov 5, 1998||Oct 22, 2002||Heidelberger Druckmaschinen Ag||Method of obtaining colorimetric values|
|US6560582||Jan 5, 2000||May 6, 2003||The United States Of America As Represented By The Secretary Of The Navy||Dynamic memory processor|
|US6594382||Nov 4, 1999||Jul 15, 2003||The United States Of America As Represented By The Secretary Of The Navy||Neural sensors|
|US6618713||Nov 4, 1999||Sep 9, 2003||The United States Of America As Represented By The Secretary Of The Navy||Neural directors|
|US6694049||Aug 17, 2000||Feb 17, 2004||The United States Of America As Represented By The Secretary Of The Navy||Multimode invariant processor|
|US6735579||Jan 5, 2000||May 11, 2004||The United States Of America As Represented By The Secretary Of The Navy||Static memory processor|
|US6768815||May 10, 2001||Jul 27, 2004||The United States Of America As Represented By The Secretary Of The Navy||Color sensor|
|US6801655||May 10, 2001||Oct 5, 2004||The United States Of America As Represented By The Secretary Of The Navy||Spatial image processor|
|1||Land, Edwin H. "An Alternative technique for the Computation of the Designator in the Retinex Theory of Color Vision," Proceedings of the National Academy of Sciences, vol. 83, 1986.|
|2||Land, Edwin H. "Experiments in Color Vision," Scientific America, vol. 200, May 5, 1959.|
|3||Zeki, S. "A Vision in the Brain," Blackwell Scientific Publishing, 1993.|
|U.S. Classification||382/162, 348/273, 358/1.9|
|International Classification||G06K9/00, G01J3/46|
|Cooperative Classification||G01J3/465, G01J3/462|
|European Classification||G01J3/46C, G01J3/46E|
|Dec 10, 2007||AS||Assignment|
Owner name: THE UNITED STATES OF AMERICA AS REPRESENTED BY THE
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:WOODALL, ROGER L.;REEL/FRAME:020223/0077
Effective date: 20010426
|Aug 2, 2011||CC||Certificate of correction|
|Sep 23, 2011||FPAY||Fee payment|
Year of fee payment: 8
|Dec 29, 2015||FPAY||Fee payment|
Year of fee payment: 12