Research Article
Multistage Optimization Using a Modified Gaussian Mixture Model in Sperm Motility Tracking
Table 2
Summary of the previous methods in multiple sperm tracking.
| Techniques | Advantages | Disadvantages |
| CASA [28, 29] | (i) Detect the sperm’s features | (i) Requires a big sample (ii) Fails in sperm collisions and proximity | JPDAF [20, 21, 22] | (i) Adaptive learning (ii) Works well with sperm collisions | (i) Computational complexity (ii) Slow process termination | HGDT [23] | (i) Combination of object segmentation and tracking (ii) Self-learning capability | (i) Fails in multiple sperm tracking. | CSR-DCF [30] | (i) Implementing of training and testing data for tracking and features extraction (ii) Usage of sperm’s features in tracking | (i) Requires large training data (ii) High computational complexity (iii) Difficult to implement | ADT [26, 31] | (i) Fast execution process (ii) Multisperm tracking | (i) Its efficiency decreases in a high dense medium. (ii) It works poorly when many sperms accumulate in a small space | DAT [27] | (i) Easy to implement (ii) Less mathematical complication | (i) Does not predict the dead and newborn sperms (ii) It does not require a testing data (frames) |
|
|