Length phenotyping with interest point detection

Plant phenotyping is the task of measuring plant attributes. We term 'length phenotyping' the task of measuring the length of a part of interest of a plant. The recent rise of low cost RGB-D sensors, and accurate deep networks, provides new opportunities for length phenotyping. In this paper we present a general technique for measuring length, based on three stages: object detection, point of interest identification, and a 3D measurement phase. We address object detection and interest point identification by training network models for each task, and use robust de-projection for the 3D measurement stage. We apply our method to two real world tasks: measuring the height of a banana tree, and measuring the length, width, and aspect ratio of banana leaves in potted plants. Our results indicate satisfactory measurement accuracy, with less than 10% deviation in all measurements. The two tasks were solved using the same pipeline with minor adaptations, indicating the general potential of the method.