Login

Join for Free!
118288 members
table of contents table of contents

This paper assessed the reliability of this simple, freely available software package …


Biology Articles » Conservation Biology » Spot the match – wildlife photo-identification using information theory » Figures

Figures
- Spot the match – wildlife photo-identification using information theory

Figures

..................................................

Figure 1

Example species with sufficient spot patterning that could be useful for automated photo-identification. Shown are (a) whale shark (Rhincodon typus – Photo © G. Taylor) indicating the reference area defined as the area encompassed by the reference points (yellow circles); (b) spotted tree frog (Hyla leucophyllata – Photo © D. Bickford); (c) northern quoll, (Dasyurus hallucatus – Photo © J. Kirwan); (d) Amazon spotted frog (Hyla punctata – Photo © D. Bickford); (e) striped blue crow (Euploea mulciber – Photo © D. Lohman); and (f) mangrove snake (Boiga dendrophilia – Photo © D. Bickford).

figure 1

..................................................

Figure 2

An individual whale shark at varying angles of yaw (A: 0°, B: 10°, C: 20°, D: 30°, E: 40°). Sequences such as this were used to assess the effect of horizontal angle on the I3S matching process.

figure 2

..................................................

Figure 3

I3S matching validation IC weights (w1). Distribution of IC weights for known matched (a) and non-matched pairs (b), and I3S matching validation evidence ratios (ER1) for known matched (c) and non-matched pairs (d) are shown.

figure 3

..................................................

Figure 4

Matching validation results. Box-and-whisker plots of (a) IC weights (w1) for known matched pairs showing images matched and not matched with I3S; (b) evidence ratios (ER1) for known matched pairs showing images matched and not matched using I3S. Central tendency (black horizontal line) indicates the median, and whiskers extend to 0.5 of the inter-quartile range

figure 4

..................................................

Figure 5

Automated versus by-eye matching results. Box-and-whisker plots of (a) IC weights (w1) for by-eye matched images that were matched and not matched using I3S; (b) Evidence ratios (ER1) for by-eye matched images that were matched and not matched using I3S. Central tendency (black horizontal line) indicates the median, and whiskers extend to 0.5 of the inter-quartile range.

figure 5

..................................................

Figure 6

Effect of angles of yaw. Box-and-whisker plots of (a) IC weights (w1) for horizontal angle categories, where images at 0° were matched against images skewed by 10°, 20°, 30° and 40°. Dotted lines show results for non-matching pairs; (b) evidence ratios (ER1) for horizontal angle categories, where images at 0° were matched against images skewed by 10°, 20°, 30° and 40°. Central tendency (black horizontal line) indicates the median, and whiskers extend to 0.5 of the inter-quartile range.

figure 6

..................................................

Figure 7

Effects of spot-pair number. (a) Relationship between complementary log-log-transformed (clog-log) I3S scores and log10-transformed number of spot pairs. The fitted line illustrates the correlation observed using a linear regression; (b) Comparison of clog-log-transformed w1 with log10-transformed number of spot pairs.

figure 7

..................................................


rating: 0.00 from 0 votes | updated on: 26 Apr 2007 | views: 4166 |

Rate article:







excellent!bad…