Does anyone have experience using map engine beta for stitching together large wooded areas (>100 acres). I have attempted to map several areas that are heavily wooded have trouble getting the images to stitch correctly every time although the quality gets reported as excellent. Individual photo quality has been verified as good too.
Not done one that big yet, but not had stitching issues, either. Can you tell us a bit more about your setup (platform, camera, height) and objectives (dsm / mosaic / NDVI)? Also, I assume you are on standard overlap on such a big mission?
I am using a phantom 3 at 380ft for these missions. The goal is to get an up to date map of what the trees look like today from the air so a forester can do on the ground representative surveys of the areas of trees that are similar age and density to try and avoid cruising the whole track of timber to estimate value. We also would like to be able to map an area that we plan to cut a portion of and then fly it post cut to measure the acreage that was actually cut. I have been using standard overlap. I considered using high overlap, but it is getting to be fall here and there are a lot of unique details in the photos where the hardwoods are turning fall colors so I am surprised stitching is a problem. I have stitched similar images in Pix4D without a problem so I am curious if I am doing something wrong with the upload.
I’ve got a 40 acre mixed deciduous woodland map on standard overlap mid summer on here which seems fine. Are you saying the issue is with the size? Do smaller acreages work ok? Can you post an example of the issue?
I did 2 maps of a fire zone here in california they worked well
http://www.dronedeploy.com/app/viewer?lng=-121.594524711&lat=36.3898368691&zoom=18.0&token=eyJhbGciOiJIUzUxMiIsInR5cCI6IkpXVCJ9.eyJ0eXBlIjoiUHVibGljVmlld2VyIiwiaWQiOiI1NjBmNGY5ZTMxZmY5YTJmNGRkNDIxM2MiLCJleHAiOjI1MzQwMjMwMDc5OX0.KWQMHIR8o0kkfxikDZpFSxXGqUVgqH7QurHwdHxgSx9h9BP9Yy_2DVzYTSMkX6yD7Z_ZgcuUrU_Ic76oHixMkw&view=562004f231ff9a04a252fb29
AND
The models came out excellent as well… https://skfb.ly/HwPQ