GENERAL STATEMENT OF THE CLASS SUBJECT MATTERThis is the generic class for apparatus and corresponding methods for the automated analysis of an image or recognition of a pattern*. Included herein are systems that transform an image for the purpose of
(a) enhancing its visual quality prior to recognition,
(b) locating and registering the image relative to a sensor or stored prototype, or reducing the amount of image data by discarding irrelevant data, and
(c) measuring significant characteristics of the image.
(1) Note. Automated document pattern* analysis or verification, which includes detection of alphanumerics, is classified in this class.
(2) Note. To be classified herein, no actual recognition or identification need be performed. It is sufficient that substantial digital image processing, such as a coding, enhancement, or transformation process, be performed on the image data for classification herein.
LINES WITH OTHER CLASSESPattern* analysis or verification limited to the intrinsic properties of a document is classified elsewhere. Documents that are analyzed or verified by information content, such as pattern*s or alphanumeric characters, are classified in this class (382). Document verification limited to a photocell system is classified elsewhere. See References to Other Classes, below.
Alphanumeric characters and other pattern*s are to be distinguished from coded indicia. Coded indicia are designed specifically to facilitate reading by machine and are not intended to be read by humans (e.g., the Universal Product Code on grocery items). Reading or sensing of coded indicia which does not include the recognition of any alphanumeric character or pattern* is classified elsewhere. However, reading or sensing of pattern*s or alphanumeric characters in combination with coded indicia is classified in this class (382). Example: Reading a credit card that contains a printed name plus a magnetic code is classified in Class 235 if only the magnetic code is read. Otherwise, if both the printed name and the magnetic code are read, classification is herein.
The images analyzed and processed herein are images that are representative of a "real" scene (such as images obtained by a camera, scanner, or image detector), including obtained images of people, places, and things, wherein the image represents the actual scene. The presentation or generation of images that are
(a) computer generated or otherwise artificial, or
(b) a combination of computer-generated images and real images is properly classified elsewhere, including for computer graphics and control of data presentation with creation or manipulation of graphic objects or text performed by a computer or processor, and operator interfaces. See References to Other Classes, below.
The specific processing of television pictures and signals, where a television system is an integral part of the system, is properly classified elsewhere. See References to Other Classes, below. When images generated by a television camera are processed, and the television system is not an integral part of the overall system, and the system is either disclosed or claimed in an environment with substantial digital image processing or in a pattern* recognition* environment, proper classification is herein.
For systems directed to the processing of a displayed image, where the processing is directed to the altering of the display image or of the display system itself, proper classification is elsewhere. See References to Other Classes, below.
The testing or measuring of distances, areas, volumes, thicknesses, or defects in objects is excluded from this class. However, where the measurements of these parameters are either disclosed or claimed in an environment with substantial digital image processing or in a pattern* recognition* environment, classification is in this class.
Image analysis* having specific and significantly claimed utility in art environments external to this class is classified in the appropriate external classes unless it is specifically excluded therefrom. For example: radar, facsimile, color facsimile, coded record sensor; and purely optical systems for image processing are all excluded from this class. See references to Other Classes, below.
Subcombinations specific to image analysis* or pattern* recognition are classified herein. |
-
100 • APPLICATIONS
-
101 • Mail processing
-
102 • ZIP code
-
103 • Target tracking or detecting
-
104 • Vehicle or traffic control (e.g., auto, bus, or train)
-
105 • License plate
-
106 • Range or distance measuring
-
107 • Motion or velocity measuring
-
108 • Surface texture or roughness measuring
-
109 • Seismic or geological sample measuring
-
110 • Animal, plant, or food inspection
-
111 • Textiles or clothing
-
112 • Document or print quality inspection (e.g., newspaper, photographs, etc.)
-
113 • Reading maps, graphs, drawings, or schematics
-
114 • Reading aids for the visually impaired
-
115 • Personnel identification (e.g., biometrics)
-
116 • Using a combination of features (e.g., signature and fingerprint)
-
117 • Using a characteristic of the eye
-
118 • Using a facial characteristic
-
119 • Using a signature
-
120 • Sensing pressure together with speed or acceleration
-
121 • Sensing pressure only
-
122 • Sensing speed or acceleration only
-
123 • Sensing geometrical properties
-
124 • Using a fingerprint
-
125 • Extracting minutia such as ridge endings and bifurcations
-
126 • With a guiding mechanism for positioning finger
-
127 • With a prism
-
128 • Biomedical applications
-
129 • DNA or RNA pattern reading
-
130 • Producing difference image (e.g., angiography)
-
131 • Tomography (e.g., CAT scanner)
-
132 • X-ray film analysis (e.g., radiography)
-
133 • Cell analysis, classification, or counting
-
134 • Blood cells
-
135 • Reading paper currency
-
136 • Reading coins
-
137 • Reading bank checks (e.g., documents bearing E-13B type characters)
-
138 • Reading monetary amount
-
139 • Reading MICR data
-
140 • Including an optical imager or reader
-
141 • Manufacturing or product inspection
-
142 • Bottle inspection
-
143 • Inspection of packaged consumer goods
-
144 • Mask inspection (e.g., semiconductor photomask)
-
145 • Inspection of semiconductor device or printed circuit board
-
146 • Measuring external leads
-
147 • Inspecting printed circuit boards
-
148 • At plural magnifications or resolutions
-
149 • Fault or defect detection
-
150 • Faulty soldering
-
151 • Alignment, registration, or position determination
-
152 • Tool, workpiece, or mechanical component inspection
-
153 • Robotics
-
154 • 3-D or stereo imaging analysis
-
155 • LEARNING SYSTEMS
-
156 • Neural networks
-
157 • Network learning techniques (e.g., back propagation)
-
158 • Network structures
-
159 • Trainable classifiers or pattern recognizers (e.g., adaline, perceptron)
-
160 • Generating a standard by statistical analysis
-
161 • Alphanumerics
-
162 • COLOR IMAGE PROCESSING
-
163 • Drop-out color in image (i.e., color to be removed)
-
164 • Image segmentation using color
-
165 • Pattern recognition or classification using color
-
166 • Compression of color images
-
167 • Color correction
-
168 • HISTOGRAM PROCESSING
-
169 • With a gray-level transformation (e.g., uniform density transformation)
-
170 • With pattern recognition or classification
-
171 • For segmenting an image
-
172 • For setting a threshold
-
173 • IMAGE SEGMENTATION
-
174 • Using projections (i.e., shadow or profile of characters)
-
175 • Separating document regions using preprinted guides or markings
-
176 • Distinguishing text from other regions
-
177 • Segmenting individual characters or words
-
178 • Separating touching or overlapping characters
-
179 • Segmenting hand-printed characters
-
180 • Region labeling (e.g., page description language)
-
181 • PATTERN RECOGNITION
-
182 • Limited to specially coded, human-readable characters
-
183 • Characters formed entirely of parallel bars (e.g., CMC-7)
-
184 • With separate timing or alignment marks
-
185 • Ideographic characters (e.g., Japanese or Chinese)
-
186 • Unconstrained handwriting (e.g., cursive)
-
187 • On-line recognition of handwritten characters
-
188 • Writing on ordinary surface (i.e., electronics are in pen)
-
189 • With a display
-
190 • Feature extraction
-
191 • Multispectral features (e.g., frequency, phase)
-
192 • Feature counting
-
193 • Counting intersections of scanning lines with pattern
-
194 • Counting individual pixels or pixel patterns
-
195 • Local or regional features
-
196 • Slice codes
-
197 • Directional codes and vectors (e.g., Freeman chains, compasslike codes)
-
198 • Extracted from alphanumeric characters
-
199 • Pattern boundary and edge measurements
-
200 • Measurements made on alphanumeric characters
-
201 • Point features (e.g., spatial coordinate descriptors)
-
202 • Linear stroke analysis (e.g., limited to straight lines)
-
203 • Shape and form analysis
-
204 • Topological properties (e.g., number of holes in a pattern, connectivity, etc.)
-
205 • Local neighborhood operations (e.g., 3x3 kernel, window, or matrix operator)
-
206 • Global features (e.g., measurements on image as a whole, such as area, projections, etc.)
-
207 • Waveform analysis
-
208 • With a tapped delay line
-
209 • Template matching (e.g., specific devices that determine the best match)
-
210 • Spatial filtering (e.g., holography)
-
211 • With electrically controlled light modulator or filter
-
212 • Nonholographic optical mask or transparency
-
213 • Using both positive and negative masks or transparencies
-
214 • With a display
-
215 • Using dynamic programming or elastic templates (e.g., warping)
-
216 • At multiple image orientations or positions
-
217 • Electronic template
-
218 • Comparator
-
219 • Determining both similarities and differences
-
220 • Calculating weighted similarity or difference (e.g., don‘t-care areas)
-
221 • Counting difference pixels
-
222 • Using an Exclusive-OR gate
-
223 • Resistor matrix
-
224 • Classification
-
225 • Cluster analysis
-
226 • Sequential decision process (e.g., decision tree structure)
-
227 • With a multilevel classifier
-
228 • Statistical decision process
-
229 • Context analysis or word recognition (e.g., character string)
-
230 • Trigrams or digrams
-
231 • Checking spelling for recognition
-
232 • IMAGE COMPRESSION OR CODING
-
233 • Including details of decompression
-
234 • Parallel coding architecture
-
235 • Substantial processing of image in compressed form
-
236 • Interframe coding (e.g., difference or motion detection)
-
237 • Gray level to binary coding
-
238 • Predictive coding
-
239 • Adaptive coding (i.e., changes based upon history, activity, busyness, etc.)
-
240 • Pyramid, hierarchy, or tree structure
-
241 • Polygonal approximation
-
242 • Contour or chain coding (e.g., Bezier)
-
243 • Shape, icon, or feature-based compression
-
244 • Lossless compression
-
245 • Run-length coding
-
246 • Huffman or variable-length coding
-
247 • Arithmetic coding
-
248 • Transform coding
-
249 • Fractal
-
250 • Discrete cosine or sine transform
-
251 • Quantization
-
252 • Error diffusion or dispersion
-
253 • Vector quantization
-
254 • IMAGE ENHANCEMENT OR RESTORATION
-
255 • Focus measuring or adjusting (e.g., deblurring)
-
256 • Object boundary expansion or contraction
-
257 • Dilation or erosion (e.g., opening or closing)
-
258 • Line thinning or thickening
-
259 • Skeletonizing
-
260 • Image filter
-
261 • Adaptive filter
-
262 • Median filter
-
263 • Highpass filter (i.e., for sharpening or enhancing details)
-
264 • Lowpass filter (i.e., for blurring or smoothing)
-
265 • Recursive filter
-
266 • Edge or contour enhancement
-
267 • Minimize discontinuities in dot-matrix image data (i.e., connecting or merging the dots)
-
268 • Minimize discontinuities at boundaries of image blocks (i.e., reducing blocking effects or effects of wrap-around)
-
269 • Minimize jaggedness in edges (e.g., anti-aliasing)
-
270 • Variable threshold, gain, or slice level
-
271 • Based on the results of a count
-
272 • Based on a local average, mean, or median
-
273 • Based on peak levels
-
274 • Intensity, brightness, contrast, or shading correction
-
275 • Artifact removal or suppression (e.g., distortion correction)
-
276 • IMAGE TRANSFORMATION OR PREPROCESSING
-
277 • Transforming each dimension separately
-
278 • Correlation
-
279 • Convolution
-
280 • Fourier transform
-
281 • Walsh, Hough, or Hadamard transform
-
282 • Selecting a portion of an image
-
283 • Using a mask
-
284 • Combining image portions (e.g., portions of oversized documents)
-
285 • Mapping 2-D image onto a 3-D surface
-
286 • Measuring image properties (e.g., length, width, or area)
-
287 • Detecting alignment marks
-
288 • Determining center of gravity or moment
-
289 • Determining amount an image is rotated or skewed
-
290 • Where the image is a character, word, or text
-
291 • Determining the position of an object
-
292 • Where the object is a character, word, or text
-
293 • Changing the image coordinates
-
294 • Registering or aligning multiple images to one another
-
295 • To position or translate an image
-
296 • To rotate an image
-
297 • Rotation of image is limited to 90 degrees, 180 degrees, or 270 degrees
-
298 • To change the scale or size of an image
-
299 • Raising or lowering the image resolution (e.g., subpixel accuracy)
-
300 • Interpolation
-
301 • Where the image is an alphanumeric character
-
302 • Multilayered image transformations
-
303 • Pipeline processing
-
304 • Parallel processing
-
305 • Image storage or retrieval
-
306 • Using identification indicia on document
-
307 • General purpose image processor
-
308 • Morphological operations (i.e., local neighborhood operations)
-
309 • EDITING, ERROR CHECKING, OR CORRECTION (E.G., POSTRECOGNITION PROCESSING)
-
310 • Correcting alphanumeric recognition errors
-
311 • Including operator interaction
-
312 • IMAGE SENSING
-
313 • Hand-held
-
314 • Sensing mechanism in stylus
-
315 • Sensing mechanism in platen
-
316 • Curve tracer
-
317 • Sensor control (e.g., OCR sheet controls copier or fax)
-
318 • Multiple scanning
-
319 • Prescanning
-
320 • Magnetic
-
321 • Optical (e.g., OCR)
-
322 • Single spot
-
323 • Single line
-
324 • Full retina
-
325 • MISCELLANEOUS
|