Fruit Detection and Tracking for Apple Counting and Mapping

Results related to the research article entitled “Video-Based Fruit Detection and Tracking for Apple Counting and Mapping” [1], presented to the IEEE MetroAgriFor2023, International workshop on Metrology for Agriculture and Forestry (November 6-8, 2023).

Fruits were detected using the YOLOv5 CNN and tracked using the ByteTrack. The methodology is applied to map fruit production in an experimental apple orchard at two different dates: four weeks and one week before harvest.

Code and dataset

The code and the data set used for fruit detection and tracking can be downloaded from the following link:  KA_ZED_Video_Fruit_Counting

References

[1] Gené-Mola J, Felip-Pomés M, Net-Barnés F, Morros JR, Miranda JC, Arnó J, Asín L, Lordan J, Ruiz-Hidalgo J, Gregorio E. 2023. Video-Based Fruit Detection and Tracking for Apple Counting and Mapping. IEEE International Workshop on Metrology for Agriculture and Forestry (MetroAgriFor). Pisa, Italia, 6-8 Novembre 2023. https://doi.org/10.1109/MetroAgriFor58484.2023.10424135