A method of operating an image recognition system including providing a neural network including a plurality of input neurons, a plurality of output neurons and an interconnection weight matrix; providing a display including an indicator; initializing the indicator to an initialized state; obtaining an image of a structure; digitizing the image so as to obtain a plurality of input intensity cells and define an input object space; transforming the input object space to a feature vector including a set of n scale-, position- and rotation- invariant feature signals, where n is a positive integer not greater than the plurality of input neurons, by extracting the set of n scale-, position- and rotation-invariant feature signals from the input object space according to a set of relationships I k =∫.sub.Ω ∫I(x,y)h[k,I(x,y)]dxdy, where I k is the set of n scale-, position- and rotation-invariant feature signals, k is a series of counting numbers from 1 to n inclusive, (x,y) are the coordinates of a given cell of the plurality of input intensity cells, I(x,y) is a function of an intensity of the given cell of the plurality of input intensity cells, Ω is an area of integration of input intensity cells, and h[k,I(x,y)] is a data dependent kernel transform from a set of orthogonal functions, of I(x,y) and k; transmitting the set of n scale-, position- and rotation- invariant feature signals to the plurality of input neurons; transforming the set of n scale-, position- and rotation- invariant feature signals at the plurality of input neurons to a set of structure recognition output signals at the plurality of output neurons according to a set of relationships defined at least in part by the interconnection weight matrix of the neural network; transforming the set of structure recognition output signals to a structure classification signal; and transmitting the structure classification signal to the display so as to perceptively alter the initialized state of the indicator and display the structure recognition signal for the structure.

Method and apparatus for image recognition using invariant feature signals

Products having optically detectable straight line segments are inspected for acceptability by forming one or more one-dimensional images of the product in which properly aligned straight line segments are respectively focused to points in the image. Such parameters as the location and image intensity of these one-dimensional image points are used to determine whether or not the product is acceptable. An optical Hough transform underlies these product inspection techniques.

Methods and apparatus for optical product inspection

An apparatus for illuminating and forming an image of a specimen has a transparent specimen-supporting surface and a fiber-optic bundle divided into illumination fibers which illuminate the surface and emission fibers which are held in the same relative spatial orientation throughout their length so that they produce an image of a specimen on the surface. This image may be analyzed by an optical demultiplexer using a light gate in which groups of input pixels are associated with a single output pixel, the light gate being switchable between various light-transmission states in each of which only one of the input pixels in each groups is allowed to transmit to the output pixel, so that the light gate effectively divides a single image comprising a large number of pixels into a series of images each containing fewer pixels. This enables a single photodetector chip to produce electronic signals representing an image having a much larger number of pixels than the number of separate photodetectors on the chip.

Apparatus and methods for forming images and for optical demultiplexing

An apparatus for labeling three-dimensionally connected components in a three-dimensional image is provided. The apparatus includes a two-dimensional labeling processor which labels each connected component in two-dimensional images giving a proper component code, and a coordinate calculator which calculates coordinates of a representative point representing each connected component. The apparatus further includes a connectivity detection processor which checks connectivity of representative points in adjacent two-dimensional image planes and an editing processor which modifies the component codes already provided by the two-dimensional labeling processor using the connectivity information to obtain three-dimensionally labeled image codes.

Apparatus and method for labeling connected component in a three-dimensional image

According to the present method of positioning a semiconductor member by examining it, an optical image of the semiconductor member is focussed on a photoelectric conversion face formed by having a plurality of photoelectric conversion elements arranged thereon in the form of a matrix. By means of the photoelectric conversion elements spotlighted by said optical image, the distributions of the spotlighted elements and the remaining non-spotlighted elements are examined through scanning operation. By using the examination results, a table having said semiconductor members loaded thereon is so moved as to permit one of them to fall under the photoelectric conversion face, thereby to inspect for positioning examination of the semiconductor member the presence or absence of a defect-indicating mark affixed thereto and examine the qualification or non-qualification of the member in terms of outer dimension, in terms of the position taken after moved, and in terms of whether or not the semiconductor member can be subjected to correction of the position, or is damaged, or is cracked, or is stained or is partially oversized. Thus, according to the present method, only each qualified one of the semiconductor members can be positioned precisely in a prescribed position, and a die bonding apparatus of the invention is based on the using of the above method for purpose of positioning the semiconductor member by examining it.

Method of positioning a semiconductor member by examining it and a die bonding apparatus using the same

A means of electrically extracting optical information from an optically recognizable pattern regardless of the orientation of the pattern within the field of view and comparing the information extracted against electrical criteria so as to classify the pattern wherein the pattern image is projected onto an image slicer which utilizes fiber optics to divide the image into a plurality of slices. Electrical processing is used to assign a number to each slice which is proportional to the light flux incident upon the slice and to square the assigned number. All the squared numbers are added to generate a voltage which is memorized. The image is then rotated with respect to the slicer and subsequent voltages are generated and memorized. The memorized voltages are then compared to an electrical template to determine the classification of the pattern and its orientation within the field of view.

Richard D Brugger

Papers

Rotation translation independent feature extraction means