Data processing: measuring – calibrating – or testing – Measurement system – Orientation or position
Reexamination Certificate
2003-03-12
2004-07-20
Barlow, John (Department: 2863)
Data processing: measuring, calibrating, or testing
Measurement system
Orientation or position
C702S149000, C702S152000, C701S213000, C701S214000
Reexamination Certificate
active
06766270
ABSTRACT:
FIELD OF THE PRESENT INVENTION
The present invention relates to machine vision and more particularly to a method and system for determining an object's position and orientation in a 3 dimensional space. The system can be used to aid the object detection and identification process.
BACKGROUND OF THE PRESENT INVENTION
There is a demand for determining an object's position and orientation using machine vision. The current technology utilizes complex image processing algorithms for 2 dimensional or 3-dimensional images. Arbitrary range and orientation configurations of an object typically cause such processing algorithms to be unreliable and slow—this has limited the utilization of machine vision to very few applications where the operating environments (with respect to lighting, object's shape and orientation) are tightly controlled.
The disadvantage associated with processing only image data to extract object's position and orientation is the necessary capability to process extraordinary amounts of information. Different poses and sizes of an object and lighting conditions manifest in large variations of image signatures. Previous methods for dealing with image data included pattern matching techniques. These techniques rely upon models and correlation algorithms in order to recognize and isolate objects within image data. These pattern matching techniques are limited by their lack of robustness. Their results degrade significantly across spatial and temporal variations. They are inadequate at recognizing objects from a broad range of pose, size, and lighting variations. They require a tremendous amount of computation. Also, these techniques do not respond well to increasing number of models (or object templates) in a database needed to detect the variations within the image data.
It is also noted that the image data represent a complex set of functions and that object images are not easily detected and identified without prior knowledge of objects and background characteristics. The traditional method for object detection and identification using image data is to use a large number of reference signatures (or training sets). The disadvantage of this method is huge computational load and low accuracy.
SUMMARY OF THE PRESENT INVENTION
A main objective of a gravity-reference vision system of the present invention is to provide an effective method to combine an inertial sensor with ranging and imaging devices and establish vision processing under leveled conditions. An example system configuration of the present invention is shown in FIG.
1
. An inertial sensor which is sensitive to the gravity field can be employed to determine leveled conditions for the sensing platform—this defines a horizontal (leveled) plane where horizontal object dimensions and position can be easily estimated from a ranging device, wherein a sweep of range data is made by rotating either the ranging device or its steering beam-in-plane. Thus, processing of imaging data from an imager (which provides a vertical plane information) to extract the object's position/orientation becomes trivial when the knowledge of the previously defined viewing plane and object's estimated range/orientation in the horizontal plane is exploited. Processing of the integrated inertial sensor, ranging device and imager data provides a complete 3-dimensional position/orientation of an object.
It is a further objective of the present invention to establish a horizontal viewing of range-to-object information using an inertial sensor along with a ranging device.
It is a further objective of the present invention to provide an image processing method that employs object's position and orientation information in the horizontal (leveled) plane in the extraction of position and orientation in the vertical plane, and thus completes the 3-dimensional position/orientation determination of an object.
It is a further objective of the present invention to determine different object's surfaces as viewed along the horizontal (leveled) plane by bifurcating two different segments along the range measurements—defining one segment to the left of the minimum range point and another segment to the right of the minimum range point.
It is a further objective of the present invention to accommodate the tilt angles as defined by the ground slope and provide a generic definition of a leveled condition by transforming the gravity constant into respective XYZ components along the body-fixed axes.
REFERENCES:
patent: 5112126 (1992-05-01), Graebner
patent: 6427122 (2002-07-01), Lin
American GNC Corporation
Barlow John
Chan Raymond Y.
David and Raymond Patent Group
Vo Hien
LandOfFree
Gravity-reference vision 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 Gravity-reference vision system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Gravity-reference vision system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3194725