Drawing on CRIM’s expertise in machine learning and artificial intelligence, Effigis Géo-Solutions has enhanced ScanSol, its platform for predicting soil texture from satellite images.
Specializing in processing geospatial data, the service company Effigis Géo-Solutions offers innovative tools to help modernize industrial sectors. Effigis called on CRIM to improve its ScanSol system, used to predict soil texture from satellite images. CRIM provided its expertise in remote sensing and machine learning to improve the existing platform.
We asked CRIM to give us a hand, because they are experts in the field.
— Claire Gosselin, Effigis Géo-Solutions
Making precision agriculture accessible
The maps currently used for understanding soil texture are problematic because they are not detailed enough to allow for optimization of each agricultural area.
This is where the ScanSol system comes in : it can predict soil texture from data extracted from satellite images. These predictions provide farmers with information that enables them to optimize their yields, for example by choosing the most suitable soil for growing certain varieties or the level of irrigation expected for each plot in their field.
CRIM’s mission was to increase the performance of the ScanSol system by adding, sorting and analysing a large amount of new data. CRIM experts also sought to automate the estimation of the percentage of various substances covering each soil zone: sand, silt, clay, or vegetation.
The ScanSol platform is already commercialized, but thanks to these recent improvements, Effigis Géo-Solutions will be able to attract more clients from the Quebec and Canadian agricultural sector.