Following the just mentioned theme, ABACO started an important test with SIN (The Italian Information System Company for the Agriculture Development) which resulted in the design and creation of a software solution to provide assistance in the procedure of control of the quality of land cover and maximum eligible areas determined in an LPIS data model.
Thanks to this collaboration, QC LPIS (Quality Control Land Parcel Identification System) is born, a new highly-engineered project that is able to take advantage of the several digital sources available (validated technical cartographies, aerial orthophotos, DTM, DSM, VHR, etc.), nonetheless of specific algorithms for spatial and spectral image analysis.
This process consists of four main steps:
- topological analysis of the "Object Regions" (OR), to ensure that the provided land cover is topologically correct; - reference vector data tests, to provide a reliable result, since the reference data used are certified by third parties; - recognizing algorithm tests, to provide a "may be" result, in the sense that image processing algorithms are likely to detect some particular land characteristic; - likelihood, to give a "probabilistic" answer: no evident signature of a particular land cover, although the test provides a likelihood of land cover. |
It concludes then with a shortlist of territorial areas, described in the census database, that the software labels as "for checking", since they are not compatible with what it was interpreted by the algorithms.