General
      Back to home
      1. Pix4D Documentation
      2. PIX4Dcatch
      3. General

      How to use Automatic Tag Detection and Manual GCP Marking - PIX4Dcatch RTK

      Coded Tags can now be used directly in PIX4Dcatch, which are detected and marked automatically by the application. Users only need to load the Point Collection that contains the GCPs, allowing for PIX4Dcatch projects to achieve higher absolute accuracies leveraged by GCPs without the need for tedious manual marking.

      IN THIS ARTICLE

      Download Tags for Auto-Detection
      Best Practices
      Automatic Tag Detection
      Manual Marking of GCPs and Tags

      Download Tags for Auto-Detection

      Tags are uniquely coded targets that the application can automatically detect in images. Tags are ideal and commonly used for marking GCPs on the field. Tags can be downloaded here: A4, US Letter.

      Tip: Tags can be printed on a sticker and glued to a firm plate (e.g. thin plywood). This makes the tags rigid and durable. Some print shops also provide services for printing the tags on vinyl or plastic, which makes them more durable and water resistant.

      Best Practices

      Placing of Tags/GCPs

      • The Tags/GCPs should be placed on the ground at a sufficient distance from each other.
      • The images should not be blurry and the Tags/GCPs should be clean.
      • The Tags/GCP should not be obstructed by any object above them nor by shadows.
        Example: The screenshots below show potentially problematic Tags/GCPs for detection.

        Pix4D_AutoGCP_obstructed_1.jpg

        Pix4D_AutoGCP_obstructed_3.jpg

        Pix4D_AutoGCP_obstructed_2.jpg

      • The Tag/GCPs must be large enough in order to be visible in the images.
        Tip: We recommend printing the tags on A4 paper size.
      • Good image overlap makes the automatic Tag Detection results more reliable, with each tag visible in multiple images.
      • If unused Tags are visible in the images, automatic Tag Detection can be misleading.

      Number and Distribution of Tags/GCPs

      The Tags/GCPs should be evenly distributed in the area of interest. Imagine the area as a large table and the Tags/GCPs as the legs that will support it. If all the "legs" are placed at the same location of the "table", then it will tilt. If the "legs" are homogeneously spread, then the "table" will be stable. Additionally, it is also recommended to place one Tag/GCP in the center of the area to further increase the quality of the reconstruction.

      Important:
      • It is mandatory to use a minimum of three Tags/GCPs to be able to use them in the photogrammetry process. To anchor your project with better absolute accuracy, it is recommended to use 5 to 10 Tags/GCPs, which is also sufficient for large projects. More Tags/GCPs do not contribute significantly to increasing accuracy and avoiding drifts.
      • The Tags/GCPs should be placed evenly on the landscape to minimize the error in scale and orientation.
      • Do not place the Tags/GCPs exactly at the edges of the area, as they will only be visible in a few images.
      • The minimum number of Tags/GCPs required for a corridor project depends on the same factors as many other types of projects, including the relative accuracy of the image geolocation, the amount of image overlap, and the length and width of the corridor.
      • We recommend that you distribute your Tags/GCPs so that they are offset from one another, regardless of the number of Tags/GCPs that you include in your project. You can consider including a pair of GCPs at each end of the corridor in addition to the set of offset Tags/GCPs that you collected along the length of the corridor.
      Example:

      corridor_mapping_1.png

      Automatic Tag Detection

      Tags can be automatically detected by PIX4Dcatch. The only requirement is to select the Point Collection before starting the capturing process.

      Important: The Point Collection's output coordinate reference system has to be a projected CRS.

      Automatic Tag Detection - Summary

      • Display targets on the scene
      • Measure the location of the targets
      • Turn on the Tag Detection option
      • Select your Point Collection
      • Start capture
      • Process project with GCPs

      Automatic Tag Detection - Detailed Description

      To select the Point Collection, first visit the PIX4Dcatch_ToolsMenu.png Tools at the bottom and select Tag Detection.

      PIX4Dcatch_autotag02_a.png

      Select the Point Collection to use for the automatic tag detection.

      GCP_marking_import_collection_2.png

      Important: Matching the tag is based on the last digits of the point names in the Point Collection. Please make sure that the point names end in the same number as the one visible on the tag. Other formats are also acceptable as long as the point ends in a number. For more information: How to measure single points(GCPs) - PIX4Dcatch RTK
      Examples:
      • PointName_1 corresponds to tag 01
      • PointName_2 corresponds to tag 02
      Important: Currently, Pix4D supports 55 different tags and a maximum of 55 point suffixes can be mapped.

      Start capturing a project in which the tags are visible. Below the Signal Indicator dialog, the total number of successfully marked tags and the number of point entries in the Point Collection is displayed. If a tag is detected in the image, it will be marked with the corresponding name on white background. If the tag is detected in at least 3 images, the tag label will turn green with a black background. To use a Tag/GCP for processing, it has to be detected in at least 3 images.

      PIX4Dcatch_autotag03_a.png PIX4Dcatch_autotag03_b.png

      During capturing, at least 3 of the marked Tags/GCPs need to be displayed with green labels. Once the project capturing is finished, a capture report will be displayed with the scanned scene and the marked tags.

      PIX4Dcatch_autotag04.png

      After capturing, the Point Collection and detected marks can be exported by clicking on Export Points & Marks. These detected marks replace the tedious manual marking and can be used as automatically marked GCPs or as MTPs for further processing in photogrammetry software.

      autoTag_export_points_marks.PNG

      In Pix4Dmatic, the marks can be imported and used as MTPs to improve the 3D reconstruction (1st screenshot below). Alternatively, the GCPs and automatic marks can be imported to improve the positioning accuracy of the 3D reconstruction (2nd screenshot below). To process a project with Tags, please visit Processing projects - PIX4Dcatch.

      Tip: For a more detailed review of the automatic tag detection, please open the Project; on the bottom part, select GCP_marking_mark_points.png Mark Points. The tags imported to the Point Collection, the list of images, and a map of the capturing path will be displayed. By selecting a tag and an image, the marking accuracy can be validated. The map helps to identify the images where the selected tag was detected. To add or remove markings, please follow the procedure described in the next section: Manual marking of Tags and GCPs.

      Importing Marks in Pix4Dmatic

      Pix4DMatic_AutoTags_MTP_workflow.png

      Importing GCPs and Marks in Pix4Dmatic

      Pix4DMatic_AutoTags_GCP_workflow.png

      Manual Marking of GCPs and Tags

      GCPs and Tags can be marked manually in PIX4Dcatch to have fully automated processing in PIX4Dcloud.

      1. file_blk.jpg Open the Project management dialog.

        PIX4Dcatch_projects.jpg

      2. Select the project.

        GCP_marking_grey.png

      3. GCP_management.png Open the Point Collection selection dialog you want to use as GCPs.

        GCP_marking_import_collection.png

        Important: The Point Collection has to be measured in a projected coordinate reference system.
      4. Close the dialog; the GCP icon will turn blue and the selected Point Collection will be displayed. Now, the Mark points is also available.

        PIX4Dcatch_GCP_marking_blue.png

      5. GCP_marking_mark_points.png Select Mark points.
      6. Select the GCP from the list to mark.

        PIX4Dcatch_GCP_marking_selection_b.png

      7. Mark GCP on an image with a single tap. Hold tapping a bit longer to open the magnifier for more accurate marking. To expand an image or delete a marking, tap the 3 dots at the top right of each image.

        PIX4Dcatch_GCP_marking_simple_b.png PIX4Dcatch_GCP_marking_magnifier_b.png

      8. Mark the GCP in at least 2 images.
      9. GCPs will be shown on the map; marked images will be displayed in green on the map.
      10. Tap Save.
      11. Select and mark additional GCPs for the project as before.
      Tip: For maximum absolute accuracy, mark GCPs in as many images as possible and also use the maximum zoom level to be able to mark the GCP at the pixel size level.

      To process a project with GCPs, please visit Processing projects - PIX4Dcatch.