Poor results from a mapping mission with significant elevation changes and dense vegetation.

PIX4Dmapper
1

Hello everyone,

I am using an Autel Evo II Pro to fly mapping missions of conservation properties. Most of the maps for these properties have turned out fine despite a majority of the properties consisting of dense vegetation, however, I am really struggling with one of our properties. We have mapped it twice now and both times the mosaic and point cloud have had significant errors. After the first attempt, I tried troubleshooting by reading the quality report help page and reviewing some of the suggestions on different posts in this community. For the second flight, we flew a grid mission at 395 feet AGL with a front overlap of 85% and a side overlap of 80% and the gimbal pitch at 80 degrees. At this point, I cannot fly the drone any higher and can’t increase the overlap much more to fix the problem. There is a fairly significant change in elevation across this property and I am wondering if that is the cause of our issues, but the areas with uncalibrated images are not the areas of highest elevation (despite what the DSM is indicating).

I know many people have had this issue - has anyone found a good way to fix it? Normally we just refly a property with different settings when we run into issues, but I don’t think that will fix the problem this time.

Map
Map1632×725 180 KB
Omara2021Oct25_report.pdf (2.8 MB)

3

Additionally, when viewing the ray cloud it looks like some of the images were taken at different elevations/angles. Does anyone know why this is happening?

PointCloud
PointCloud1920×1323 142 KB

4

Hi natalie,
Mapping heavily dense vegetation often times causes processing problems. Vegetation inherently has complex geometry that tends to move in the breeze. This makes it difficult for Pix4Dmapper to generate keypoints and as a result some images might not get calibrated. Processing with lower Keypoint Image Scale can lead to a higher number of calibrated images than the default original keypoint image scale. Try processing at 1/2 or 1/4 image scale.

5

Hi Mike,

Having the same problems as Natalie, would one expect any different results using Pix4dcloud, than pix4dmapper?

6

Hi pfrancis,
PIX4Dcloud and PIX4Dmapper use the same processing engine. However Cloud has limited processing options as compared to Mapper. Cloud only lets you select the predefined templates for processings. It is fine if everything goes well but if there is an issue such as heavy vegetation and you need to reduce the image scale then you will need to use Mapper.

7

hey Mike, can you add in Manual tie points to uncalibrated images? If yes what is the best way to do this?

8

Unfortunately it is the same problem. Cloud just doesn’t give you the option to utilize MTPs. However you can very easily add MTPs in Mapper which often helps to calibrate images.