Agricultural pests are responsible for millions of dollars in crop losses and management costs every year. In order to implement optimal site-specific treatments and reduce control costs, new methods to accurately monitor and assess pest damage need to be investigated. In this paper we explore the combination of unmanned aerial vehicles (UAV), remote sensing and machine learning techniques as a promising technology to address this challenge. The deployment of UAVs as a sensor platform is a rapidly growing field of study for biosecurity and precision agriculture applications. In this experiment, a data collection campaign is performed over a sorghum crop severely damaged by white grubs (Coleoptera: Scarabaeidae). The larvae of these scarab beetles feed on the roots of plants, which in turn impairs root exploration of the soil profile. In the field, crop health status could be classified according to three levels: bare soil where plants were decimated, transition zones of reduced plant density and healthy canopy areas. In this study, we describe the UAV platform deployed to collect high-resolution RGB imagery as well as the image processing pipeline implemented to create an orthoimage. An unsupervised machine learning approach is formulated in order to create a meaningful partition of the image into each of the crop levels. The aim of the approach is to simplify the image analysis step by minimizing user input requirements and avoiding the manual data labeling necessary in supervised learning approaches. The implemented algorithm is based on the K-means clustering algorithm. In order to control high-frequency components present in the feature space, a neighbourhood-oriented parameter is introduced by applying Gaussian convolution kernels prior to K-means. The outcome of this approach is a soft K-means algorithm similar to the EM algorithm for Gaussian mixture models. The results show the algorithm delivers decision boundaries that consistently classify the field into three clusters, one for each crop health level. The methodology presented in this paper represents a venue for further research towards automated crop damage assessments and biosecurity surveillance.
In this paper we discuss how we see the capabilities of DBMSs evolving over the next several years to meet the needs of expert data base applications. We also present some of the research thrusts which we see as important that appear to be receiving insufficient attention in the research community.
Adequate information modeling in non-standard application areas (e.g. engineering applications such as CAD/CAM, VLSI design or knowledgebased applications) requires the abstraction concepts of classification, aggregation, generalization, and association. The Molecule-Atom Data model (MAD) designed for the effective support of such an information model is justified and described with its essential properties and features. MAD offers dynamic object definition and object handling, based on direct and symmetric management of network structures and recursiveness. These generic mechanisms can be used to map the above mentioned abstraction concepts in a straight-forward manner. Thus, the mapping of a wide variety of semantic and object-oriented modeling constructs, including complex objects with shared subobjects, becomes feasible. All these concepts are illustrated by means of some vivid examples taken from the areas of CAD/CAM and knowledge-based applications.
object in an object-oriented data model
KIVIEW is an object-oriented browsing interface to databases. The underlying internal model is a semantic network, defined via a collection of triplet facts which are governed by a small set of integrity requirements. KIVIEW's only ezternal structure is a display, called uiew, that presents all the information that is pertinent to a given object. KIVIEW features two kinds of activity, called nawgation and manipulation. Navigation is an exploration activity wherein the user repeatedly displays new views of objects that are referenced in current views. To speed up repetitive navigational tasks, which otherwise could become very tedious, viewa inay be eynchronized: the user sets up several viewa, linked in cr tree-like structure, so that when the information displayed in the root view is modified, the contents of the other views change automatically. Manipulation is a processing activity wherein the user operates on existing views to create new virtual views. Manipulation is interleaved with navigation to assemble information located while browsing into reeulte of browsing sessions. Using convenient graphic tools, these two activities are integrated into a single tool, which is flexible and effective, yet simple to use. This work waa supported in part by the European Communitiee under ESPRIT project 1117; the work of Motro was also supported by NSF Grant No. IRE8609912 and by an Amoco Foundation Engineering Faculty Grant; the work of Tarantino waa also supported by ENIDATA. 107 108 EXPERT DATABASE SYSTEMS
