Table 1. Techniques and datasets used for WBC segmentation.
The WBC segmentation techniques dataset used for WBC segmentation, along with the relevant features and image enhancement techniques commonly applied before the WBCs segmentation process.
| Work | Database | Segmentation | Features | Enhancement | Classifier |
|---|---|---|---|---|---|
| Sharma & Buksh (2019) | ALL-IDB1 108 images | Hybrid segmentation using ALL-DC model | Morphological feature | Firefly optimization algorithm (FOA) based lighting technique | Fuzzy based CNN |
| Al-jaboriy et al. (2019) | ALL-IDB1 108 images | ROI-based segmentation using local pixel information | 4-moment statistical features | Nil | ANN |
| Abdulhay et al. (2018) | 100 microscope images | ROI and edge detection | Local binary patterns and texture features | Median filter and adaptive weighted median filter, | Support vector machine |
| Shahin et al. (2018) | BS_DB3, ALL_DB1, 2 108 and 260 images |
Otsu’s thresholding | Nil | Smoothing procedure with average filter | Modified watershed transform |
| Quiñones et al. (2018) | 12 blood smear images |
Zak algorithm | Eccentricity and area | HSV color space | Nil |
| Cao, Liu & Song (2018) | 203 and 70853 drown themselves | SWAM&IVFS, fuzzy divergence based | Texture | RGB and HIS color space | Nil |
| Miao & Xiao (2018) | 100 manually collected images | Marker-controlled watershed | Nil | Histogram base technique | Marker-controlled watershed |
| Nikitaev et al. (2018) | 1,018 cells element |
Modified watersheds/ distance transformation | Nil | luminance histogram analysis | Supervised learning method |
| Shirazi et al. (2016c) | Nil | Snake algorithm, Gram-Schmidt orthogonalization | Nil | curvelet transform, wiener filter | Nil |
| Di Ruberto, Loddo & Putzu (2016) | ALL-IDB, 108 images SMC-IDB 367 images IUMS-IDB 195 images | Segmentation Via SVM | Texture feature | NNS (nearest neighbor search) with Euclidean distance | Nil |
| Liu et al. (2015) | 306 images from hospital | Nucleus mark watershed, Mean shift clustering | Color feature | RGB and HIS color space | Nil |
| Tosta et al. (2015) | 367 images | Neighbourhood valley-emphasis | Nil | A median filter and linear contrast stretching | Nil |
| Shirazi et al. (2016b) | ALL-IDB1 108 images | Thresholding and mathematical morphology | Geometrical and texture | Wiener filter and curvelet transform | Back-propagation neural network |
| Alomari et al. (2014) | 100 microscopy images | Thresholding | Color feature | Nil | Nil |