The resulting trajectory fragments then needs to be fused in order to reconstruct the complete movement of each animal. When an occlusion takes place, ToxId cuts the tracked trajectories of the animals involved in the occlusion. ToxId can be included into the workflow of any online tracking algorithm and can detect and handle occlusions of multiple animals in an efficient way. The software, the user manual and the documentation are available at. In addition, it can also measure the time an organism spends near aquaria or terrarium walls. ToxId is implemented in the free tracking software ToxTrac 5, which allows the user to extract locomotor activity such as average speed, acceleration and distance traveled per time unit. ToxId requires no training or complex configuration steps, it does not use features or characteristic fingerprint-like maps, and it requires significantly less memory than other algorithms 6, 18. ToxId identifies animals in 97% of the frames and achieves this using the intensity histogram and the Hu-moments of detected animals. We show that ToxId achieves the same accuracy as the best state-of-the-art algorithm, also when considering error propagation. ToxId can handle a large variety of animals and it does not use future or past frames, thus it can be used as a post processing stage in real-time applications. To overcome the mentioned limitations - animal specific algorithms, slow algorithms that operate only offline, and low accuracy - we developed a new online algorithm called ToxId. In these situations the use of an online tracking is required 19. Thus, offline tracking is generally slow and cannot be applied in real-time, streaming applications, or other situations where only the current frame can be accessed. These techniques are computationally and memory expensive and require access to past and future frames (offline tracking). Other approaches to reduce occlusion problems rely on pattern recognition, matching specific texture maps 6 or using convolutional neural networks 18 to identify the animals. Nevertheless these methods can only be applied for animals geometrically compatible with the used model. Additionally, some authors discuss the use of features such as face properties 16 or bilateral symmetry 17. To improve detection and tracking, some techniques use a specific model of the animal body based on the head shape 10, 11, the body geometry 12, 13, 14 or the symmetry axis 15. This technique adds complexity to the experimental setup and dramatically increases the amount of data generated. Other methods rely on improving the detection by using several cameras with different perspectives 8, 9. These solutions, however, may be invasive and are not applicable to the vast majority of animals. To address the occlusion problem, some techniques tag the organisms with a visual marker to preserve their identity 7.
0 kommentar(er)
