Image analysis – Image transformation or preprocessing – Mapping 2-d image onto a 3-d surface
Reexamination Certificate
1998-10-22
2001-10-16
Boudreau, Leo (Department: 2721)
Image analysis
Image transformation or preprocessing
Mapping 2-d image onto a 3-d surface
C382S141000, C382S154000, C345S440000
Reexamination Certificate
active
06304680
ABSTRACT:
BACKGROUND OF THE INVENTION
This invention relates generally to process monitoring systems, and, more particularly to a high resolution, high accuracy monitoring system capable of 3D process monitoring with 2D video and a precise angular aiming system.
Manufacturing errors are expensive. An error in manufacturing results in costs relating to raw materials, machine time, resolution time, and rework or scrap disposal.
The cost of a manufacturing error is lowest when the error is detected and corrected at the earliest possible time in the process. Each process which occurs after a manufacturing error has occurred adds costs to the product. The flawed product will eventually cause a manufacturing disruption or become scrap. A dimensioning error which is not detected when the dimension is created can allow the repeated production of parts which will not function properly.
It has been said that quality cannot be “inspected into” parts. Quality is the result of each and every process occurring correctly. By monitoring dimensional quality “in process” (as the process is occurring) problems can be detected and resolved at the earliest possible time and at the lowest possible cost. The detection of developing problems, during process deterioration but before unacceptable processing occurs, lowers costs and improves the efficiency of a manufacturing system. Costs are reduced by resolving the problem before scrap is created or downstream disruptions occur.
A single fixed camera has a resolution which generally does not exceed 640 by 480 pixels. Single camera inspection has been limited to 2D inspection of small areas (tens of square feet). Large inspection fields are supported by mounting the camera on a gantry. The gantry is generally 2D, and the cost and complexity associated with precision over large areas rapidly becomes prohibitive.
Another method of expanding the field size of a single camera is to aim the camera through a pair of XY galvanometric mirrors. While this increases the field size for the single camera, accuracy and precise focus are limited. Inspection cameras aimed through mirrors are limited to 2D applications.
Another drawback of single camera systems (stand alone, on gantries or aimed with mirrors) is the complexity and unreliability of the video processing task for the data read in from the camera.
It is therefore an object of this invention to provide 3D process monitoring with a 2D video system.
It is a further object of this invention to position a camera's viewing field in small regions of interest which may be occur anywhere over a large area.
It is still a further object of this invention to control inspection parameters in processing the data captured by a video camera.
It is another object of this invention to generate 2D views of 3D characteristics for the purpose of limiting regions of interest in a video camera inspection system.
SUMMARY OF THE INVENTION
The objects set forth above as well as further and other objects and advantages of the present invention are achieved by the embodiments of the invention described hereinbelow.
The present invention uses transforms to generate 2D views of 3D characteristics as defined in 3D CAD data. The invention (high resolution, high accuracy 3D process monitoring with a 2D video camera and a precise angular aiming system) allows automated inspection to occur faster, more reliably, more accurately, and over larger areas and volumes (3D space) than existing inspection systems.
3D dimensional inspection requires two cameras and typically has limited field size (tens of square feet except in large, expensive systems). The invention expands this field size and volume by an order of magnitude (to hundreds of square feet with comparable accuracy) without requiring the complexity of multiple cameras and equipment to move the cameras.
CAD data is used to define most assembly and dimensional manufacturing processes, both manual and automated. The ability to use the CAD definition of a 2D or 3D dimensional process for automatic in-process quality monitoring can yield cost reductions related to the frequency and impact of errors occurring in the manufacturing process. The invention allows automatic documentation and verification in answer to the question: “are things where they are supposed to be?”
Inspection solutions in use today search for edges and calculate where things have been found. Such systems often find edges which are unrelated to the current inspection task. The invention first calculates exactly what it is looking for and where it should be. The invention then looks only precisely where a process characteristic should be and determines whether or not the characteristic is present. This simplifies detection tasks and reduces vision system errors. Detection of spurious or contradicting characteristics is eliminated because the vision system operates only in the precise region of interest of the tolerance band.
Mapping of 3D definition onto the single 2D camera allows the vision system to recognize geometries which would be potentially confusing to a 2D system.
The system of the present invention maintains a 3D model which is accurately referenced to the part being inspected.
This 3D model mapped onto the 2D camera field enhances accurate process characteristic recognition.
The savings resulting from the application of the invention fall into several categories:
The invention provides a less expensive inspection system than current products with higher performance inspection capabilities. The higher performance characteristics include: larger field, higher resolution and accuracy, and more reliable characteristic detection.
Manufacturing costs are reduced through the application of the invention. Costs relating to scrap, rework, and process stoppage are reduced through the earliest possible detection and resolution of process errors.
The invention can provide inexpensive, traceable documentation of accurate processing for each part produced. This electronic documentation is automatically generated and made available for storage. Analysis of this data for Statistical Process Control (SPC) can yield improvements in production efficiency.
The process documentation automatically generated for each inspection can take the form of dimensional information or pictures. This data is generated by a computer and can be stored on inexpensive computer media such as disk or tape. This documentation has value in proving to customers that each part is known to be correct through a non-subjective, traceable to a standard, measurement system. This is a performance advantage which has economic value in competitive production situations. Traceable documentation is becoming a standard with the advent of ISO 9000 procedures.
The invention also has value in its ability to improve the performance of vision software which is used to decipher characteristics from frames of video. Data created by the invention is used to improve detectability of inspection characteristics. This data also reduces the amount of processing required to complete an inspection task.
The invention saves time and effort by providing an automated process which can operate where inspection is not currently occurring or to replace many inspection processes which are currently manual and subjective. Manual inspection requires the presence of an inspector when a process has been completed. Inspectors cannot be everywhere at once. Waiting for a human inspector to arrive and complete the inspection task is a significant overhead cost.
For a better understanding of the present invention, together with other and further objects thereof, reference is made to the accompanying drawings and detailed description and its scope will be pointed out in the appended claims.
REFERENCES:
patent: 4700225 (1987-10-01), Hara et al.
patent: 4845558 (1989-07-01), Tsai et al.
patent: 4989082 (1991-01-01), Hopkins
patent: 5091963 (1992-02-01), Litt et al.
patent: 5129009 (1992-07-01), Lebeau
patent: 5157735 (1992-10-01), Maeda et al.
patent: 5384523 (1995-01-01), Masuda
patent: 5388
Blake Scott W.
Fenning Frederick S.
Kayes Lucille M.
Lapshin Ilya
Assembly Guidance Systems, Inc.
Boudreau Leo
Cohen Jerry
Erlich Jacob N.
Kayes Lucille M.
LandOfFree
High resolution, high accuracy process monitoring system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with High resolution, high accuracy process monitoring system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and High resolution, high accuracy process monitoring system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2617736