grand_summary: Summary statistics for evaluated crowns

In weecology/NeonTreeEvaluation_package: Evaluate Individual Tree Segmentation Algorithms against Benchmark Data

Summary statistics for evaluated crowns

Description

summary_statistics returns a standardized summary of plot evaluations. The mean recall and precision can be calculated either for the entire dataset, or for each of the 22 geographic sites.

Usage

grand_summary(results, threshold = 0.4)

Arguments

results	A data frame of matched predictions and ground truth returned from evaluate_plot
threshold	Float. Intersection-over-union threshold to consider a prediction a true positive.
by_site	Logical. Should average recall and precision be calculated for each geographic site separately? Defaults to FALSE, such that a single summary is returned for all sites.
calc_plot_level	Calculate plot-level statistic comparing the total predictions to the total crowns. This only makes sense for the image-annotated crowns, and not the field-annotated crowns.

Value

A names list of overall mean precision and recall, by site, and plot level summary

weecology/NeonTreeEvaluation_package documentation built on Aug. 27, 2024, 10:53 a.m.