Research Article

A VidEo-Based Intelligent Recognition and Decision System for the Phacoemulsification Cataract Surgery

Table 4

The performance of probe tracking on the testing set of the standard Tacking-Learning-Detection (TLD) and the adaptive TLD methods.

Video Precision of adaptive TLD (%) Precision of TLD (%)

2 84.5 61.2
3 91.4 15.5
4 98.2 82.0
5 76.2 64.4