Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11106936-B2
Publication Date
2021-08-31
Expiration Date
Abstract
Point clouds of objects are compared and matched using logical arrays based on the point clouds. The point clouds are azimuth aligned and translation aligned. The point clouds are converted into logical arrays for ease of processing. Then the logical arrays are compared (e.g. using the AND function and counting matches between the two logical arrays). The comparison is done at various quantization levels to determine which quantization level is likely to give the best object comparison result. Then the object comparison is made. More than two objects may be compared and the best match found.
Core Innovation
The invention compares point cloud data for objects by selecting a first analog point cloud representing a first object and a second analog point cloud representing a second object, then azimuth aligning the first and second point clouds. After azimuth alignment, the invention translates and translation aligns the first and second point clouds to improve correspondence. The aligned point clouds are then voxelized and quantized and used to generate regularly spaced logical arrays at a first resolution and at a second resolution.
At each resolution, the invention compares the logical arrays generated from the first and second point clouds and generates a comparison response value. The comparison relies on logical array comparison operations, producing a quantified comparison response that reflects how well the point clouds match at that resolution. The invention also generates regularly spaced logical arrays at a selected quantization level corresponding to selected spatial accuracy, and a comparison response curve is produced to support selecting the appropriate response.
The invention selects one of the comparison response values and generates a comparison result for the first object and the second object based upon the selected response value. In described implementations, alignment uses center of mass or center of gravity, weighted directional statistics or bump-map derived characteristic signatures, and cross-correlation to select the angular/azimuth alignment; translation alignment uses offsets that maximize logical-AND match counts. The multi-resolution point-cloud comparison is represented as a point-cloud pyramid with comparison performed across multiple resolution levels, enabling best-of-class comparison and separability.
Claims Coverage
Two independent claims are present, directed respectively to a method for comparing two point clouds and to an apparatus configured to perform the method. Across these independent claims, the inventive features center on azimuth and translation alignment, converting quantized point clouds into regularly spaced logical arrays at multiple resolutions, computing comparison response values from logical array comparisons, selecting a comparison response value, and outputting a comparison result.
Azimuth alignment followed by translation alignment
Azimuth aligning the first and second point clouds, and translation aligning the first and second point clouds after the azimuth alignment.
Quantization and multi-resolution logical arrays
Quantizing the first and second point clouds and generating multiple regularly spaced logical arrays at multiple resolutions based upon the first and second point clouds.
Logical array comparison to generate resolution response values
Comparing the logical arrays based upon the first point cloud to logical arrays based upon the second point cloud at various resolutions and generating a comparison response value at the various resolutions.
Selecting a comparison response value for final comparison result
Selecting one of the comparison response values and generating a comparison result for the first object and the second object based upon the selected response value.
The independent claims cover both a point-cloud comparison method and an apparatus implementing it, with the core inventive structure being azimuth/translation alignment, quantization into regularly spaced logical arrays at multiple resolutions, Boolean logical array comparison to produce response values, and selection of a response value to generate the final comparison result.
Stated Advantages
Efficient and compact point-cloud comparison using logical arrays/bit arrays rather than larger float or octree structures.
Separability of objects based on comparison response behavior across quantization levels.
Ability to compare and identify a best match based on the selected comparison response value.
Documented Applications
Vehicle object comparison using CAD vehicle point clouds and SAR-derived vehicle point clouds, including exemplar matching and best-of-class comparison.
Interested in licensing this patent?