Assessment of Mixed Sward Using Context Sensitive Convolutional Neural Networks

Abstract

Breeding higher yielding forage species is limited by current manual harvesting and visual scoring techniques used for measuring or estimation of biomass. Automation and remote sensing for high throughput phenotyping has been used in recent years as a viable solution to this bottleneck. Here, we focus on using RGB imaging and deep learning for white clover (Trifolium repens L.) and perennial ryegrass (Lolium perenne L.) yield estimation in a mixed sward. We present a new convolutional neural network (CNN) architecture designed for semantic segmentation of dense pasture and canopies with high occlusion to which we have named the local context network (LC-Net). On our testing data set we obtain a mean accuracy of 95.4% and a mean intersection over union of 81.3%, outperforming other methods we have found in the literature for segmenting clover from ryegrass. Comparing the clover/vegetation fraction for visual coverage and harvested dry-matter however showed little improvement from the segmentation accuracy gains. Further gains in biomass estimation accuracy may be achievable through combining RGB with complimentary information such as volumetric data from other sensors, which will form the basis of our future work.

Identification of Weeds Based on Hyperspectral Imaging and Machine Learning

“…Considering that each pot was scanned individually, the presence of shadows was not a major issue. For cases where the scanning is performed in the field, it is possible to use a different illumination system to reduce the presence of shadows ( Bateman et al, 2020 ).…”

Image‐based high‐throughput phenotyping for the estimation of persistence of perennial ryegrass (Lolium perenneL.)—A review

“…Manual methods have a higher possibility of under‐ or over‐estimating species, depending on the size of the species or the size of the investigation area, which will in turn lead to having poor accuracy of species composition estimation. However, image sensors and advanced image processing algorithms have a great potential to assess pasture composition under field conditions (Bateman et al., 2020; Skovsen et al., 2017).…”

Robust Species Distribution Mapping of Crop Mixtures Using Color Images and Convolutional Neural Networks

“…Consistent with previous research in the field, the biomass composition of the mixed crop is estimated based on the visible canopy in a top-down camera view [2,3,7,8,10,19]. Setup as a semantic segmentation task, where every pixel is discretely classified, every biomass image is described by 9 × 10 6 plant species classifications.…”

“…Therefore, automated and robust methods for estimating clover proportion in mixtures are needed for example to adjust fertilization levels at 'close to real-time'. Although the technique has been shown to work with deep learning based methods [2,3], there is a need to verify and enhance the stability of model predictions across different growth conditions, camera systems, and mixtures of species and varieties.…”

“…Bateman et al [2] proposed an improved model architecture for semantic segmentation, named LC-Net (Local Context Network). Based on a VGG16 model [13] backbone, an encoder-decoder architecture was designed to fuse standard feature maps with contextual feature maps of lower resolution.…”