Research Article
U-DAVIS-Deep Learning Based Arm Venous Image Segmentation Technique for Venipuncture
Table 2
Comparison of different augmentation methods on the basis of their PSNR, IoU, and dice scores.
| | PSNR | IoU | Dice | Training with: | Mean | Std | Min | Max | Mean | Std | Min | Max | Mean | Std | Min | Max |
| No augmentations | 0.435 | 0.146 | 0.068 | 0.678 | 0.682 | 0.01 | 0.552 | 0.961 | 0.397 | 0.11 | 0.067 | 0.517 | AHE | 0.582 | 0.165 | 0.045 | 0.827 | 0.788 | 0.007 | 0.659 | 0.973 | 0.48 | 0.141 | 0.045 | 0.708 | CLAHE | 0.624 | 0.165 | 0.097 | 0.886 | 0.79 | 0.005 | 0.718 | 0.98 | 0.545 | 0.146 | 0.097 | 0.799 | Both AHE & CLAHE | 0.751 | 0.155 | 0.117 | 0.93 | 0.893 | 0.004 | 0.821 | 0.996 | 0.685 | 0.149 | 0.117 | 0.871 |
|
|