Camera calibration parameter about pyk4a

[lpasselin]

This could be documented better. Both in our wrapper and on the official SDK.

See functions save_calibration_jsonand load_calibration_json

pyk4a/pyk4a/pyk4a.py

Line 74 in 6d1355a

def save_calibration_json(self, path: Any):

pyk4a/pyk4a/pyk4a.py

Line 78 in 6d1355a

def load_calibration_json(self, path: Any):

You should use save_calibration_json first to have a template.
An example can be found here: https://github.com/etiennedub/pyk4a/blob/6d1355a1ee994fddcabb48501586b058671c9443/tests/assets/calibration.json

[zhaocheng-hub]

可以更好地记录下来。 无论是在我们的包装中还是在官方SDK中。

查看功能 save_calibration_json和 load_calibration_json

pyk4a/pyk4a/pyk4a.py

Line 74 in 6d1355a

def save_calibration_json(self, path: Any):

pyk4a/pyk4a/pyk4a.py

Line 78 in 6d1355a

def load_calibration_json(self, path: Any):

您应该 使用 save_calibration_json首先 一个模板。
可以在这里找到一个示例： : https //github.com/etiennedub/pyk4a/blob/6d1355a1ee994fddcabb48501586b058671c9443/tests/assets/calibration.json

@lpasselin
I try to change some data manually, and try to use k4a. load_ calibration_ json to read the data of "JSON" in the path I specified. And k4a.calibration_raw is printed with print There is no change in the attribute of k4a.calibration_raw.

In addition, through k4a.save_calibration_json saves "JSON" data, it is effective. It changes the original "json" data (but the style becomes only one line of code instead of the style of the template you gave me).

I will k4a.load_calibration_json in the main function, I found that he successfully printed my modified "json", but this method still can't change k4a.calibration_raw

Maybe my usage is wrong. Can you give me an example?

[shagren]

You can use high level api: create Calibration from string and then use it for image transformation or coordinates translation

calibration = Calibration.from_raw(json_string, DepthMode.NFOV_UNBINNED, ColorResolution.RES_1080P)
transormed_image =  depth_image_to_color_camera(depth_image, calibration) 
color_x, color_y, color_z  = calibration.depth_to_color_3d((depth_x, depth_y, depth_z)) # 3d coordinates from Depth camera space to Color camera space

[zhaocheng-hub]

您可以使用高级api：根据字符串创建校准，然后将其用于图像转换或坐标转换

calibration = Calibration.from_raw(json_string, DepthMode.NFOV_UNBINNED, ColorResolution.RES_1080P)
transormed_image =  depth_image_to_color_camera(depth_image, calibration) 
color_x, color_y, color_z  = calibration.depth_to_color_3d((depth_x, depth_y, depth_z)) # 3d coordinates from Depth camera space to Color camera space

Thank you for your reply， this method is effective.