UWB Radar for the detection of fungal infestation in vines
Fraunhofer FHR aims to improve the early detection of stem diseases in vines with a new UWB radar sensor. The sensor, along with other sensors that are currently available, will be integrated on the mobile multisensor platform of Televitis, an R&D group at the University of La Rioja and the Institute of Grapevine and Wine Sciences (ICVV) in Spain. Televitis fosters the research of new technologies for applications in precision viticulture and develops robot platforms for the regular monitoring of wine-growing regions.
In the last 30 years, stem disorders have developed to become one of the most destructive vine diseases and they now represent a growing threat in wine-growing regions around the world. Some studies suggest that the annual economic damage resulting from the replacement of dead vines amounts to approximately 1.5 billion euro.
Stem diseases are threatening the economic existence of many winegrowers. Their pathogens attack the vital organs of the plants and sooner or later provoke their death. The diseases are caused by fungi, the life cycle and epidemiology of which are all very similar. The diseases are not evident at first and their symptoms develop very slowly over several years with the result that they are very difficult to detect. The fungal spores normally access the plant via wounds that occur during the pruning process. The pathogens can also penetrate the plant via other mechanically inflicted wounds or via damage caused by frost. As they grow, they gradually decompose the wood and cause the plant to die slowly. The fruit bodies in the dead wood and their spores are released by water, scattered by the wind and can now infect new wounds. The symptoms normally take the form of sectional or central necrosis in the wood which is recognizable by distinctive brown zones and signs of rotting. The leaves show discoloration and begin to wither, sometimes very quickly. This acute disease can not be treated. Hence, the infected plants have to be removed and burned to hopefully prevent the disease from spreading.
Early diagnosis with radar
External radar sensors can acquire information about the plant's interior. They can therefore detect the disease at an early stage and make a distinction between healthy and infected plants in the vineyard. In an earlier project, Fraunhofer FHR developed a radar sensor that is capable of monitoring crops, including an estimation of the water content and biomass of the plants. The scientists adapted this sensor for the detection of diseases in vine stems and conducted initial tests in a preliminary study. To this end, the University of La Rioja provided the scientists with samples of freshly cut vines in different stages of the stem disease.
The project aims to provide a mobile platform for the regular monitoring of wine-growing regions. This platform should be capable of evaluating information from optical, multispectral, hyperspectral and NDVI sensors.
So as to be in a position to adjust the geometry of the data recording of the mobile TELEVITIS platform also within the framework of initial laboratory tests, the FHR engineers used a robot arm which can be moved along six axes. This facilitated the testing of stripmap and spotlight recording modes in one- and two-dimensional scanning rasters. With the flexible robot arm, it was also possible to exploit polarization diversity and test other properties that are based on polarimetry. On completion of the promising laboratory tests, the scientists used their radar sensor to carry out the first tests on living plants in several vineyards in Logroño, Spain in November 2017.
3D-analyses and computer simulation optimize sensor performance
Fraunhofer FHR is following different research approaches to improve the UWB radar sensor so that stem diseases in vines can be detected at an early stage and contained so that they can not spread across the vineyard. These include denser recording grids based on circular SAR approximation values at different heights as well as imaging techniques using inverse modeling. The three-dimensional angle of sight created in this manner should be able to quantitatively and spatially reconstruct the constitutive parameters inside the stems. Using the photogrammetric reconstruction from the optical camera on the platform, it will also be possible to create 3D models of each vine. These, in turn, can be included in the SAR processing so as to further improve the inverse modeling.
In addition, the scientists want to create computer simulations of the course of the disease inside the vine stem. To this end, a large number of samples in different stages of the disease will have to be analyzed. This will lead to a better understanding of the course of the disease and, with exact simulations, the sensitivity of the radar sensor can also be increased to ensure that this major viticultural problem can be detected as early as possible.