Table 3.
Application of 3D reconstruction techniques for seed phenotype digitization.
| Optical sensor | Pros and cons | Seed | Application | Method and performance | Ref |
|---|---|---|---|---|---|
| 3D laser scanner | High-precision and high-resolution Not limited by lighting environment Cumbersome to operateTime-consuming for data collection |
Rice | 3D geometric characteristic modeling | Geomagic Studio | [122] |
| Surface shape feature measurement | Geomagic Studio, accuracy > 98 % | [125] | |||
| Maize | 3D geometric characteristic modeling | – | [126] | ||
| Legume | 3D modeling and trait measurement | Visual Studio | [121] | ||
| Grape | Variety discrimination | cluster analysis and PCA | [123] | ||
| MVS | Convenient image acquisition operation High color reproduction Time-consuming for calculationNot suitable for texture-lacked objects |
Maize | 3D geometric characteristic modeling | Visual Studio, Matlab and Python, relative error = 0.1 % | [120] |
| Arabidopsis, rapeseed, barley | 3D modeling and trait measurement | Matlab, R2 = 0.492 ∼ 0.994 | [124] |