Computer vision-based yield-to-picking area mapping for horticultural product

A system comprising:

• a harvesting implement comprising a conveyor belt through which a harvested horticultural product moves from a particular picking area of a field to a harvesting bin;
• an imaging apparatus coupled to the harvest implement configured to capture digital video data comprising frames that represent a view of the harvested horticultural product as the harvested horticultural product moves through the conveyor belt during a harvesting operation;
• one or more location sensors coupled to the harvesting implement configured to produce digital geographic location data identifying a geographic location of the harvesting implement when the digital video data was captured;
• one or more processors;
• and
• a memory storing instructions which, when executed by the one or more processors, cause performance of:
• inputting, to a machine-learned model trained to recognize a horticultural product in digital imagery, the captured digital video data;
• outputting, using the machine-learned model, annotated video data, the annotated video data comprising digital annotation data identifying individual instances of the horticultural product detected in frames of the digital video data by the machine-learned model, the digital geographic location data, and digital timestamp data comprising timestamps associated with the frames;
• using the annotated video data, computing quantitative data comprising particular counts of the individual instances of the horticultural product associated with particular timestamp data;
• using the digital timestamp data, mapping the quantitative data to the digital geographic location data to produce a digital yield map;
• and
• causing display of the digital yield map on a field manager computing device.