Fig 1.
FAST corner extraction diagram.
Fig 2.
Binary descriptor of the combination of gray size and gray value difference information.
Fig 3.
Illustrates the comprehensive workflow of the ALGD-ORB algorithm.
Table 1.
Details of each algorithm.
Fig 4.
Schematic diagram of five-directional segmentation.
Fig 5.
Optical images from Oxford dataset.
(a) Bikes (Blur) (b) Trees (Blur) (c) Graf (Viewpoint) (d) Wall (Viewpoint) (e) Bark (Zoom & Rotation) (f) Boat (Zoom & Rotation) (g) Leuven (Light) (h) UBC (JPEG Compression).
Fig 6.
Illustrates the distribution of feature points extracted by different algorithms.
(a) SIFT, (b) SURF, (c) BRISK, (d) FREAK, (e) LATCH, (f) LDB, (g) ORB, (h) Mur-ORB, (i) LF-Net, (j) D2-Net, (k) R2D2, and (l) ALGD-ORB, as demonstrated on the Leuven group image from the Oxford dataset.
Table 2.
Average feature point distribution uniformity across image groups in the Oxford dataset.
Fig 7.
Illustrates the distribution of feature points extracted by different algorithms.
(a) SIFT, (b) SURF, (c) BRISK, (d) FREAK, (e) LATCH, (f) LDB, (g) ORB, (h) Mur-ORB, (i) LF-Net, (j) D2-Net, (k) R2D2, and (l) ALGD-ORB, as demonstrated on the Leuven group image from the Oxford dataset.
Fig 8.
Matching precision results on the Oxford dataset.
(a) Bikes (blur). (b) Trees (blur). (c) Graf (viewpoint). (d) Wall (viewpoint). (e) Bark (zoom and rotation). (f) Boat (zoom and rotation). (g) Leuven (light). (h) Ubc (JPEG compression).
Table 3.
Mean precision for image groups based on the Oxford dataset.
Fig 9.
Mean RMSE analysis on the Oxford dataset.
Table 4.
Feature points are extracted from images with various scenes and lighting conditions.
The mean RMSE of each group is shown in units of pixels.
Fig 10.
Statistical analysis of mean operation time on the oxford dataset.
Table 5.
Shows the mean operation time of each algorithm on Oxford dataset in units of seconds.