# Number and distribution of GCPs in Corridor Mapping .

**URL:** <https://community.pix4d.com/t/number-and-distribution-of-gcps-in-corridor-mapping/5081>\
**Category:** PIX4Dmapper Questions/Troubleshooting\
**Created:** [April 18, 2016, 7:50am UTC](https://community.pix4d.com/t/number-and-distribution-of-gcps-in-corridor-mapping/5081 "2016-04-18T07:50:14Z")\
**Posts on this page:** 7\
**Page:** 1

<div class="post-metadata">

**Author:** ![Bikram\_Gangwar1](https://avatars.discourse-cdn.com/v4/letter/b/57b2e6/32.png) [@Bikram\_Gangwar1](https://community.pix4d.com/u/Bikram_Gangwar1)\
**Post date:** [April 18, 2016, 7:50am UTC](https://community.pix4d.com/t/number-and-distribution-of-gcps-in-corridor-mapping/5081/1 "2016-04-18T07:50:14Z")

</div>

I have to survey a 3 mile (5 KM) road using an UAV. I have gone through following links&nbsp;

[https://support.pix4d.com/hc/en-us/articles/202559299](https://support.pix4d.com/hc/en-us/articles/202559299%20)

[https://support.pix4d.com/hc/en-us/articles/202557459#label6](https://support.pix4d.com/hc/en-us/articles/202557459#label6)

and understood that there should be around 8-10 GCPs, but the distance of length of flight is not clear for 8-10 GCPs requirement. What should be the distribution of GCPs, whether it should be in zig zag ( kind of random) or it should be linear in pairs ?

The road is surrounded by the vegetation, and the UAV will be flying at the height of 100 m with&nbsp;5.2-26 mm camera.

---

<div class="post-metadata">

**Author:** ![Ignatius\_Krogh](https://avatars.discourse-cdn.com/v4/letter/i/6de8d8/32.png) [@Ignatius\_Krogh](https://community.pix4d.com/u/Ignatius_Krogh)\
**Post date:** [April 19, 2016, 11:38am UTC](https://community.pix4d.com/t/number-and-distribution-of-gcps-in-corridor-mapping/5081/2 "2016-04-19T11:38:05Z")

</div>

the software will make a more accurate surface if you have equally spaced points, but also, points that reflect the ‘highs’ and ‘lows’ of the road. I would suggest putting points on either side of the road, but also on the middle of the road (if the road drops or rises relative to the natural surface.) I.e like a 5 points pattern.

---

<div class="post-metadata">

**Author:** ![Bikram\_Gangwar1](https://avatars.discourse-cdn.com/v4/letter/b/57b2e6/32.png) [@Bikram\_Gangwar1](https://community.pix4d.com/u/Bikram_Gangwar1)\
**Post date:** [April 20, 2016, 9:44am UTC](https://community.pix4d.com/t/number-and-distribution-of-gcps-in-corridor-mapping/5081/3 "2016-04-20T09:44:39Z")

</div>

Thanks for the reply ! Can you also clarify that, for what corridor length 8-10 GCPs are required ? And how many GCPs will be sufficient for 3 miles&nbsp;corridor, if we take 5 point pattern ?

Thanks !

---

<div class="post-metadata">

**Author:** ![Ignatius\_Krogh](https://avatars.discourse-cdn.com/v4/letter/i/6de8d8/32.png) [@Ignatius\_Krogh](https://community.pix4d.com/u/Ignatius_Krogh)\
**Post date:** [April 20, 2016, 12:09pm UTC](https://community.pix4d.com/t/number-and-distribution-of-gcps-in-corridor-mapping/5081/4 "2016-04-20T12:09:07Z")

</div>

I never really went on any training&nbsp;for this so I cannot give you definitive answers on what configuration will work in your site, trial and error&nbsp;will lead to experience. why are you limited to 8-10 GCP’s? that seems fine though. 100m height flight/26mm lens. what frequency do you take pictures at? how much overlap?

in any case,&nbsp;&nbsp;if it is possible, experiment with using 3, 5, 10 GCP’s (or take 10 GCP’s and then re-run the process by systematically eliminating one from the calculation and look at the quality report, if your mean error on the gcp’s are ‘low’ enough, then you know you have a good accuracy with that GCP configuration.&nbsp;

---

<div class="post-metadata">

**Author:** ![Bikram\_Gangwar1](https://avatars.discourse-cdn.com/v4/letter/b/57b2e6/32.png) [@Bikram\_Gangwar1](https://community.pix4d.com/u/Bikram_Gangwar1)\
**Post date:** [April 20, 2016, 12:56pm UTC](https://community.pix4d.com/t/number-and-distribution-of-gcps-in-corridor-mapping/5081/5 "2016-04-20T12:56:27Z")

</div>

Actually I was going through the links -

[https://support.pix4d.com/hc/en-us/articles/202559299](https://support.pix4d.com/hc/en-us/articles/202559299%20)

[https://support.pix4d.com/hc/en-us/articles/202557459#label6](https://support.pix4d.com/hc/en-us/articles/202557459#label6)

&nbsp;

and they mentioned that 8-10 GCPs will be enough. But they didn’t mention about the corridor length. So I was confuse that for how much distance they mentioned to take 8-10 GCPs.

At every 2-3 seconds I will be taking pictures and I will be keeping \>85% frontal overlap and around 75% side overlap.

---

<div class="post-metadata">

**Author:** ![Kathy\_McCready](https://avatars.discourse-cdn.com/v4/letter/k/ad7895/32.png) [@Kathy\_McCready](https://community.pix4d.com/u/Kathy_McCready)\
**Post date:** [June 27, 2017, 11:54pm UTC](https://community.pix4d.com/t/number-and-distribution-of-gcps-in-corridor-mapping/5081/6 "2017-06-27T23:54:29Z")

</div>

I am also curious about how to distribute GCPs, and how many to include, in corridor mapping after looking at the referenced support articles.

My corridor is 3 miles (5km) by 200 feet (60 meters).

I also saw that someone got an answer of having 3 GCPs/per flight ([https://community.pix4d.com/t/4950-How-many-GCP-s-#gsc.tab=0](https://community.pix4d.com/t/4950-How-many-GCP-s-#gsc.tab=0)). I am curious if that only pertains to varied terrain or should be utilized in all cases?

---

<div class="post-metadata">

**Author:** ![jgbaldwin](https://avatars.discourse-cdn.com/v4/letter/j/8e7dd6/32.png) [@jgbaldwin](https://community.pix4d.com/u/jgbaldwin)\
**Post date:** [January 8, 2020, 7:07pm UTC](https://community.pix4d.com/t/number-and-distribution-of-gcps-in-corridor-mapping/5081/7 "2020-01-08T19:07:38Z")

</div>

Bikram,

Have you learned anything about GCP distribution since your posting? I am doing a 1.8 mi corridor flight and would appreciate any advice.
