# Can this camera output the point cloud of specific location in the image

**URL:** <https://community.stereolabs.com/t/can-this-camera-output-the-point-cloud-of-specific-location-in-the-image/762>\
**Category:** Uncategorized\
**Created:** [December 23, 2021, 7:31am UTC](https://community.stereolabs.com/t/can-this-camera-output-the-point-cloud-of-specific-location-in-the-image/762 "2021-12-23T07:31:57Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![ujs1514lee](https://community.stereolabs.com/user_avatar/community.stereolabs.com/ujs1514lee/32/313_2.png) [@ujs1514lee](https://community.stereolabs.com/u/ujs1514lee)\
**Post date:** [December 23, 2021, 7:31am UTC](https://community.stereolabs.com/t/can-this-camera-output-the-point-cloud-of-specific-location-in-the-image/762/1 "2021-12-23T07:31:57Z")

</div>

Here is my idea：  
I want to use python + YOLO +zed 2i to get the position of the objected being detected. Can I use depth\_sensing.py to get the point cloud and coordinate of specific location in the image? For example, object inside the YOLO detection box, can I get the point cloud information and coordinate information inside the box and ignore the outside information?  
Besides, I want to know, when I use Python API sl.Camera.retrieve\_measure(), how can I use GPU to speed up. I can not find type=sl.MEM.GPU, there is only type=sl.MEM.CPU, but my computer definitely has a GPU.  
Thanks, looking for your reply!

---

<div class="post-metadata">

**Author:** ![Myzhar](https://community.stereolabs.com/user_avatar/community.stereolabs.com/myzhar/32/11_2.png) [@Myzhar](https://community.stereolabs.com/u/Myzhar)\
**Post date:** [December 23, 2021, 8:27am UTC](https://community.stereolabs.com/t/can-this-camera-output-the-point-cloud-of-specific-location-in-the-image/762/2 "2021-12-23T08:27:00Z")

</div>

Hi @ujs1514lee  
I start from the last question. Currently it is not possible to use the `retrieve_measure` to store data in GPU memory with the Python language, only C++ allows this:  
[https://www.stereolabs.com/docs/api/python/classpyzed\_1\_1sl\_1\_1Camera.html#ae387f1b1c5a1454f2e2812c77267cf60](https://www.stereolabs.com/docs/api/python/classpyzed_1_1sl_1_1Camera.html#ae387f1b1c5a1454f2e2812c77267cf60)

Regarding the first question, at the moment we do not have API functions that allow to extract a region of a point cloud contained in a 3D bounding box.

PS We could add this feature in one of the next release of the SDK.