Anders Krogh Mortensen

Weight prediction of broiler chickens using 3D computer vision

Fully-automatic 3D camera-based weighing system for broiler chickens.System tested in a commercial production scenario during 20days.3D depth images from the Kinect sensor. In modern broiler houses, the broilers are traditionally weighed using automatic electronic platform weighers that the broilers have to visit voluntarily. Heavy broilers may avoid the weigher. Camera-based weighing systems have the potential of weighing a wider variety of broilers that would avoid a platform weigher which may also include ill birds. In the current study, a fully-automatic 3D camera-based weighing system for broilers have been developed and evaluated in a commercial production environment. Specifically, a low-cost 3D camera (Kinect) that directly returned a depth image was employed. The camera was robust to the changing light conditions of the broiler house as it contained its own infrared light source.A newly developed image processing algorithm is proposed. The algorithm first segmented the image with a range-based watershed algorithm, then extracted twelve different weight descriptors and, finally, predicted the individual broiler weights using a Bayesian Artificial Neural Network. Four other models for weight prediction were also evaluated.The system were tested in a commercial broiler house with 48,000 broilers (Ross 308) during the last 20days of the breeding period. A traditional platform weigher was used to estimate the reference weights. An average relative mean error of 7.8% between the predicted weights and the reference weights is achieved on a separate test set with 83 broilers in approximately 13,000 manually annotated images. The errors were generally larger in the end of the rearing period as the broiler density increased. The absolute errors were in the range of 20-100g in the first half of the period and 50-250g in the last half. The system could be the stepping stone for a wide variety of additional camera-based measurements in the commercial broiler pen, such as activity analysis and health alerts.

Dicotyledon Weed Quantification Algorithm for Selective Herbicide Application in Maize Crops

The stricter legislation within the European Union for the regulation of herbicides that are prone to leaching causes a greater economic burden on the agricultural industry through taxation. Owing to the increased economic burden, research in reducing herbicide usage has been prompted. High-resolution images from digital cameras support the studying of plant characteristics. These images can also be utilized to analyze shape and texture characteristics for weed identification. Instead of detecting weed patches, weed density can be estimated at a sub-patch level, through which even the identification of a single plant is possible. The aim of this study is to adapt the monocot and dicot coverage ratio vision (MoDiCoVi) algorithm to estimate dicotyledon leaf cover, perform grid spraying in real time, and present initial results in terms of potential herbicide savings in maize. The authors designed and executed an automated, large-scale field trial supported by the Armadillo autonomous tool carrier robot. The field trial consisted of 299 maize plots. Half of the plots (parcels) were planned with additional seeded weeds; the other half were planned with naturally occurring weeds. The in-situ evaluation showed that, compared to conventional broadcast spraying, the proposed method can reduce herbicide usage by 65% without measurable loss in biological effect.

Segmentation of lettuce in coloured 3D point clouds for fresh weight estimation

Abstract Monitoring the health and yield of crops during production is an important, but labour intensive component of commercial agriculture, especially in high value crop such as lettuce. This article proposes a novel method for segmenting lettuce in coloured 3D point clouds and estimating the fresh weight. The proposed segmentation method operates by clustering points into leaves and then evaluating their affiliation to a lettuce of interest. From the segmented lettuce point clouds, the volume, surface area, leaf cover area and height predictors are extracted and correlated to the fresh weight. The proposed segmentation and yield estimation methods are evaluated on Cos and Iceberg lettuce point clouds generated from images collected by an agricultural robot in an outdoor field experiment. The results demonstrate that the proposed segmentation method is able to successfully isolate lettuce (F1-score = 0.88–0.91). Analysis of the segmented lettuce models show that the calculated surface areas correlate strongly with measured fresh weight (R2 = 0.84–0.94). Not only does this validate the segmentation method, it allows an accurate estimate of the lettuce fresh weight (RMSE = 27–50 g) to be produced non-destructively.

Estimation of the Botanical Composition of Clover-Grass Leys from RGB Images Using Data Simulation and Fully Convolutional Neural Networks

Optimal fertilization of clover-grass fields relies on knowledge of the clover and grass fractions. This study shows how knowledge can be obtained by analyzing images collected in fields automatically. A fully convolutional neural network was trained to create a pixel-wise classification of clover, grass, and weeds in red, green, and blue (RGB) images of clover-grass mixtures. The estimated clover fractions of the dry matter from the images were found to be highly correlated with the real clover fractions of the dry matter, making this a cheap and non-destructive way of monitoring clover-grass fields. The network was trained solely on simulated top-down images of clover-grass fields. This enables the network to distinguish clover, grass, and weed pixels in real images. The use of simulated images for training reduces the manual labor to a few hours, as compared to more than 3000 h when all the real images are annotated for training. The network was tested on images with varied clover/grass ratios and achieved an overall pixel classification accuracy of 83.4%, while estimating the dry matter clover fraction with a standard deviation of 7.8%.

Preliminary Results of Clover and Grass Coverage and Total Dry Matter Estimation in Clover-Grass Crops Using Image Analysis

The clover-grass ratio is an important factor in composing feed ratios for livestock. Cameras in the field allow the user to estimate the clover-grass ratio using image analysis; however, current methods assume the total dry matter is known. This paper presents the preliminary results of an image analysis method for non-destructively estimating the total dry matter of clover-grass. The presented method includes three steps: (1) classification of image illumination using a histogram of the difference in excess green and excess red; (2) segmentation of clover and grass using edge detection and morphology; and (3) estimation of total dry matter using grass coverage derived from the segmentation and climate parameters. The method was developed and evaluated on images captured in a clover-grass plot experiment during the spring growing season. The preliminary results are promising and show a high correlation between the image-based total dry matter estimate and the harvested dry matter ( R 2 = 0.93 ) with an RMSE of 210 kg ha − 1 .