1 Introduction
2 Principles and Methods
2.1 Related Work of XXL-CT Data Image Segmentation
2.1.1 Deep Neural Networks
2.1.2 Flood Filling Networks (FFNs)
2.2 Data Acquisitionsec: Data
2.2.1 Bulk Material
2.2.2 XXL-CT
2.3 Annotation Pipeline
2.3.1 Conventional Image Processing Chain
Dataset | Instances | Merges (not intersecting) | Splits (corrected) | Boundary errors |
---|---|---|---|---|
Marble training | 39 | 1 (0) | 1 (1) | 3 |
Marble test | 37 | 2 (0) | 2 (2) | 1 |
Corn training | 457 | 18 (2) | 1 (0) | 0 |
Corn test | 461 | 21 (2) | 2 (0) | 0 |