Corn silage evaluation is an important step in determining the quality of silage harvested from a forage harvester. Corn silage is used as fodder for cattle in dairy production and high quality silage though correct processing has an effect on milk yield and sub-optimal setting of the machinery can also lead to the quality being affected significantly.
Kernels must be sufficiently cracked for efficient starch intake by lowering the requirement for chewing during eating and ruminating. Currently, the quality of kernel processing is evaluated through means requiring separation of kernel fragments and stover (leaves and stalks) which can be time consuming, cumbersome to conduct, and prone to error.
This PhD project aims to directly measure the quality of harvested corn silage from a forage harvester with an RGB camera in non-separated samples. Deep learning with Convolutional Neural Networks are trained to efficiently and robustly detect particles in the silage relevant to quality metrics. By measuring the quality directly this allows for immediate information for the farmer on their harvesting process allowing for maximizing yield and machinery usage whilst lowering requirements on the farmer during the strenuous harvesting season.
Maize Silage Kernel Fragment Estimation Using Deep Learning-Based Object Recognition in Non-Separated Kernel/Stover RGB Images
Rasmussen, C. B. & Moeslund, T. B., 10 aug. 2019, I : Sensors. 19, 16, 3506.
This Industrial PhD is in collaboration with CLAAS E-Systems and funded by Innovation Fund Denmark under the grant number: 7038-00170B.