Using aerial images taken by drone, plane or satellite, RSIP Vision develops software for image processing and analysis in forestry to efficiently determine:

  • Forest border delineation

    Automatic detection of the forest border and its sub section borders. This capability is used to update geographic databases according to the actual forest bounding polygon.

forest border

Forest border and sub sections polygons

  • Non planted areas detection

    Non planted areas may result from:

    • Planned non planting areas – Natural reserve sections, piles or rows of debris.
    • Generated during some maintenance process – Thinned rows, skidding ways and logging areas.
    • Caused during tree growing as outcome of water stress or some disease – Typically regions within the forest without trees. Their shape is arbitrary.

Natural forest segment detection

two detected areas of missing trees

Natural forest segment detection and two detected areas of missing trees due to debris piles

Natural forest segment detection

post thinning skidding with logging dock detection

Natural forest segment detection and post thinning skidding with logging dock detection

forest regions without trees

Detection of forest regions without trees

Detection of forest regions without trees

  • Drainage issues

    Areas with dead trees may show water supply shortage. CIR and NDVI are first calculated. A polygons bounding dead trees in some location shows such a problem. Topographic information may be cross correlated to verify water stress caused by area conditions.

  • Competition

    The term competition means that non planned crops, trees and weeds are existing nearby the young planted trees. They compete on resources, mostly water. Heavy competition may “kill” the planted young trees. Some of the competition, as trees, may be leftovers from previous generation of the forest that re-grow after the harvest. The detection and handling is critical, especially in the first 3 months. RSIP Vision is using advanced algorithms to identify competition from areal images. Here, same as with re-planting optimization, a competition maintenance optimization is performed.

  • Weeds and Species detection and classification

    Weeds are some form of competition to the young trees. As mention above, RSIP Vision can detect, classify and suggest an optimal maintenance procedure to handle.

Weeds detection

Weeds detection (yellow polygons)

Weeds detection (yellow polygons)

  • Supporting RGB and CIR images for bounded objects detection

    RSIP Vision algorithms can work with any set or subset of gray and color channels.

Natural forest segment and unplanted segments

Natural forest segment and unplanted segments detected out of CIR image

Natural forest segment and unplanted segments detected out of CIR image

This automatic detection and recognition enables more accurate decision making, saving time and cost, particularly in hard to reach areas. RSIP Vision has provided customized image processing software since 1987. Our clients have benefited scores of clients. Call us and we will help you too!

Share The Story