|Previous | Next|
IN THIS ARTICLE
Import drone flight images
To import images:
1. On the left sidebar, click Import Images, Import folder or Import GeoTIFF, and navigate to the folder containing the captured images.
2. Select the images or folder and click Open.
3. In PIX4Dfields, projects with RGB, multispectral or thermal images can be created.
4. An Advanced processing settings box appears when images from a supported rig are detected. To find out more information about these settings please go Advance processing settings. If the box appears, toggle on the needed settings and click APPLY or click CANCEL to not apply them.
5. Click ADD MORE IMAGES in case the images are in more than one location.
6. (optional) If processing multispectral data, a settings icon will appear beside the START PROCESSING button.
To process the images and create an orthomosaic:
1. To begin the process click START PROCESSING.
2. The import and image stitching processes begin, the layer will read Orthomosaic (processing...).
3. The progress bar will display the messages "Importing images...", "Bundle adjustment...", and "Stitching images..." and the progress percentage showing the current process.
4. (optional) Click the STOP PROCESSING icon at the bottom right corner to stop the stitching images process.
5. Once processing is completed, bellow LAYERS click the name of the Orthomosaic or the Surface Model to display them on the satellite map interface.
Article feedback (for troubleshooting, post here instead)
How to select directtory instead of files. selecting files takes forever when selectin over 1000 files....
Hi, when you wish to select multiple images within a directory you can select one and then use Control + A on a windows or Command + A on the Mac which will select them all.
It is possible to add multiple folders of images by importing one folder of images and then click Add More Images.
I'm using a micasense RedEdge camera and Pix4dMapper for to get the ortomosaics (higher quality than in Pix4Fields). After processing, five ortomosaics are obtained (one per band).
How should I do the fusion of the five ortomosaics to generate the indexes?
I want to do this in Pix4Fields because I have the cutting and comparison tools
is there a way to predetermin the bands when importing the images? (maybe by image naming convention?)
right now i have to set the band tag individually for each image after import - not fun when working with hyperspectral imagery...
Hi @Riego and @John,
Thank you very much for your questions and feedback on our article. Please note that for better maintenance we highly recommend our users to post their questions related to processing or troubleshooting on our community so other users can contribute/benefit from the answer.
@Riego depending on which solution you are using to process your imagery the ortomosaics are not recommended to be used when extracting indices but reflectance maps. The first one is color balanced which does not serve the purpose of preserving the true reflectance value of the pixels.
For visualization purpose in case, you want to make a composite out of your bands you can use QGIS.
@John when using hyperspectral imagery if EXIF of the images is correctly written then Pix4D Desktop will automatically recognize the number of the bands and further on you will only have to put the weights. See more see: How to set the band weights.
At the moment this functionality is not available in Pix4dFields.
I am looking to use Pix4D Fields but only have a single lens NIR senterra camera. Does the software support that camera?
Currently we are unable to support this specific camera due to the way it images are acquired and how our algorithms interpret the image data to produce the orthomosaic. We are working to support Senterra cameras in the new future. We are due to publish a supported camera database which will show which cameras are / are not currently supported.
I just downloaded the trial and tried to process my first field but after selecting all the images and then attempting to process them, the program crashes every time. Please help.
Can you let us know which type of camera you are using? There is a bug in v.1.2.0 which in rare instances leads to a crash when processing images on a non-supported camera. This issue has been resolved to show an error message with v.1.3.0 which will be available tomorrow.
If however your camera is not supported, we are working to support more models as quickly as possible and can advise on this once we know the type used.
Kind regards, Sam
I am using a Parrott Sequoia camera and I have processed images from this camera through your software in the past so I'm not sure where the problem lies. I even tried un-installing and re-installing pix4dfields and with the same result.
I just downloaded the latest update and now it is taking my images and did stitch them together. The problem now is that the stitch did not do a good job (I have attached a screenshot) and I can't find where I enter in my calibration images.
Thank you for updating to the latest version, this issue may be data specific so I would like for us to have a chance to look more in detail. Can you please raise a support ticket by using the button at the top of the page called Contact Support.
We look forward to your response,
Please consider opening troubleshooting topics on our community and not as article comments. This will help other users benefiting from the information as the article comments are not searchable.
Hello, when I am uploading the images to Pix4DFields from my computer, the import beggins but when it reaches 100% an error sign pops un saying "An error ocurred while stitching the images". How can I fix this?
This message usually relates to processing of images from an unsupported camera. We are releasing a fix (v.1.4) in the coming days which will resolve the majority of camera support issues when it exists in our camera database.
In the meantime, please try processing using Pix4Dfields cloud, this offers an alternative processing pipeline. If your issue persists or you would like more information specific to your camera you may raise a personal support ticket.
Hi, I'm David
I am using the Multispectral P4 whose camera has 5 multispectral bands: https://www.dji.com/p4-multispectral.
When I upload the images, the following error appears:
Processing does not end
The format of the images are as follows:
This means all the captures are not there. We give the same warning for sequoia or micasense cameras when this happens. One capture consists of all bands of a single image and these captures are assigned to one rig instance to calibrate the rig successfully. If some images are missing from a capture this will be discarded. In case more than 5% of captures are incomplete then the software does not further proceed with the calibration. In DJI P4 multispectral, the names are of the images are 000, 001, 002, 003, 004, 005 where the first one is RGB and the consecutive ones are each band image. It also might be the capture UIUD tag is missing: The UUID tag is used to assign images to a particular capture.
This seems to be hardware/firmware issue. You can reach out to DJI and they should be able to give you more information on how to troubleshoot, so from the next time all the bands are captured for each image and all the exif tags are correctly written.
I have been conducting very detailed notes on testing for 2 days and have not been able to get a single successful map processed. The stitching issues very from hexagon patterns to wave patterns. I started testing at 400ft because i figured it would be the easiest altitude to generate a map at. i started off with 400ft at 75% overlap at multiple speeds and had no success, I moved on to try 400ft at 80% overlap in different speeds and still had no success. One map in each overlap came close to good but not great or perfect. I attached a file of my testing sheet and screen shots of the stitching issues i am experiencing. There are three different maps i added to this post, they go from worst to best but none are perfect. Below is my workflow and all the other necessary information is in my testing sheet. Someone please help. I purchased the Micasense red edge MX after a horrible experience with slantrange and sentera. I am trying to make the transition over from Drone Deploy to pix4D but cannot do so until i have consistent data. I have 3 days left on my free trial and can not commit to the annual purchase if i cant get a successful map.
1. Set up drone
2. Automatic QR code capture
3. fly mission
4. process data with reflectance panel
We are in contact with you by email. Please share with us the dataset and log file, as I mentioned to you. We will check the raw images and troubleshoot the issue. As soon as we find out the issue, we will share the answer here.
Thank you very much for your patient.