Can we do some particular function by using zed2?

byby886 · July 28, 2023, 1:11am

In the case of 3d Human skeleton extraction (D: Pyroch_Learning zed-sdk master zed-sdk master body tracking body tracking python), it is completed by calling sdk. Whether we can work in the following way: we design a two-dimensional Human skeleton extraction network to get the key points, and then obtain the coordinates of the key points extracted by ourselves through the depth module. If possible, how should it be implemented and which functions are involved

SL-PY · July 28, 2023, 6:46am

Hi,

you should be able to yse something close to what we show in the depth sensing tutorial:

github.com

stereolabs/zed-sdk/blob/master/tutorials/tutorial 3 - depth sensing/python/depth_sensing.py#L70


      
          zed.retrieve_image(image, sl.VIEW.LEFT)
          # Retrieve depth map. Depth is aligned on the left image
          zed.retrieve_measure(depth, sl.MEASURE.DEPTH)
          # Retrieve colored point cloud. Point cloud is aligned on the left image.
          zed.retrieve_measure(point_cloud, sl.MEASURE.XYZRGBA)
          
          
# Get and print distance value in mm at the center of the image
          # We measure the distance camera - object using Euclidean distance
          x = round(image.get_width() / 2)
          y = round(image.get_height() / 2)
          err, point_cloud_value = point_cloud.get_value(x, y)
          
          
distance = math.sqrt(point_cloud_value[0] * point_cloud_value[0] +
                               point_cloud_value[1] * point_cloud_value[1] +
                               point_cloud_value[2] * point_cloud_value[2])
          
          
point_cloud_np = point_cloud.get_data()
          point_cloud_np.dot(tr_np)
          
          
if not np.isnan(distance) and not np.isinf(distance):
              print("Distance to Camera at ({}, {}) (image center): {:1.3} m".format(x, y, distance), end="\r")

replace the point cloud by the depth and set you keypoint coordinates as x and y and you should have what you want.

byby886 · July 29, 2023, 1:56am

Hi,thank you!
There is still a question.By zed2,we get a image from left view and right view,and finally the we only get one point cloud.I mean how can we connect the point of left or right view with the point in point cloud?

Myzhar · July 29, 2023, 9:32am

The data at index (u,v) of the point cloud in the sl::Mat corresponds to the data at index (u,v) of the sl::Mat with left color pixels.

byby886 · July 29, 2023, 12:41pm

Hi,
Can I understand it in the following way ?For example, I have a point ( 100,200 ) in the left view, and I can obtain the three-dimensional coordinates of the point from the point cloud by the way
point3D = point _ cloud.get _ value ( 100,200 )

alassagne · July 31, 2023, 3:14pm

Yes, you understood correctly