Automatic segmentation of trees in dynamic outdoor environments
This work addresses segmentation challenges for orchard and vineyard automation, but it is incremental as it builds on existing superpixel and color-based techniques.
The paper tackled the problem of segmenting trees in dynamic outdoor environments with uncontrolled illumination by using superpixels to identify background material and integrating color distribution to optimize segmentation parameters, resulting in a method suitable for tree reconstruction and apple flower detection applications.
Segmentation in dynamic outdoor environments can be difficult when the illumination levels and other aspects of the scene cannot be controlled. Specifically in orchard and vineyard automation contexts, a background material is often used to shield a camera's field of view from other rows of crops. In this paper, we describe a method that uses superpixels to determine low texture regions of the image that correspond to the background material, and then show how this information can be integrated with the color distribution of the image to compute optimal segmentation parameters to segment objects of interest. Quantitative and qualitative experiments demonstrate the suitability of this approach for dynamic outdoor environments, specifically for tree reconstruction and apple flower detection applications.