Visual inspection procedure

From Wikicap - European Commission

Go up to the main update page

If a manifest change is not or not all that manifest, visual inspection can often determine if a land change anomaly does indeed represent a true non-conformity. It offers a quick and simple stepwise method for inspecting LPIS reference parcels by searching for the not so obvious occurrence of a manifest change. There are three possible outcomes of this workflow:

  • The parcel has been stable and there is no need for update.
  • The parcel has indeed changed and further update is needed.
  • The change cannot be unambiguously confirmed or discarded by visual inspection and the parcel can be subjected to a subsequent congruency test.

To be able to perform this workflow correctly, the operator must be familiar with the specific ancillary data and metadata and can access these during the inspection:

  • the life cycle of the vegetation
  • the boundary delineation precision of the LPIS
  • an indication of the scene view angle or approximate camera position
  • an estimation of distance and area on the working scale

As indicated above, this visual inspection procedure can "mutatis mutandis" also be applied for application during the OTSC, measuring an agricultural parcel that wholly occupies one reference parcel.

The process of visual inspection is illustrated in Figure 13.


Figure 13: Visual inspection to document anomaly

Steps in this visual inspection process are:

  1. Prepare all data needed for visual inspection, i.e. documentation of the anomalies, relevant orthoimage, reference parcel geometries from the current LPIS, etc.
  2. If there is no problem with viewing the whole RP perimeter on the orthoimage, one may proceed. Else, keep the anomaly open as it is not possible to observe/measure the reference parcel on the given imagery. Process the anomaly appropriately in due time.
  3. If the observed displacement (RP vs orthoimage) of four, well identifiable ground level perimeter points in each of the four cardinal directions (N-E-S-W) is above 1.25 m, one may consider to perform an optional congruency test if the overall condition prevents to discard absence of change on sight. Else, continue.
  4. Check for the presence of manifest changes inside the reference parcel as described in the appropriate workflow (low level use case).
  5. If manifest changes have been found or the congruency test has failed, indicate in the documentation that the anomaly is confirmed and mark the anomaly “in progress”. Else record that anomaly as not confirmed.
  6. Perform a spatial cross-check with the neighbouring reference parcels by visualizing their inter-relations from geometries. Analyse the extent of the change affected, and determine how many reference parcels are involved in the anomaly.
  7. Indicate in the anomaly documentation, the unique ID number of the reference parcel and, if appropriate, the unique ID numbers of the reference parcels affected in the neighbourhood.

The above procedure offers one possible systematic sequence of checks in support of detection and reporting of changes to one or more reference parcels. There is absolutely no need to document individual steps. MS may operate an alternative procedure that leads to similar findings.



Figure 14: discovering the need for a congruency test in course of visual inspection

The figure above provides a case illustration of this decision thread:

  1. the overall situation makes sense
  2. the parcel perimeter can be visualized
  3. the perimeter is still valid
  4. no constructions appear,
  5. no conversion into non-agricultural land occurs,
  6. no part of the boundary is subject to displacements (the tree base is visible).
  7. the displacement of the corners is larger than 1.25: "YES" and a congruency test can provide a final verdict on any remaining doubts

Go up to the main update page