【Ori_Knowledge Graph】Wrong_dest_meta_and_numpy2numpy caused by wrong defaultValue meta

[xp1632]

While the 3d_image_slicing and edge_perserve smoothing VP Node works separately:

---

-We found the Knowledge Graph part, numpy2numpy give wrong dimension order of image

• the source image and the dest image are both (256,256), grayscale, 0-255,RGB
• should be exactly the same

---

-However, we get :

---

### - Reason:

• The conversion called automatically also feed wrong destination metadata :

• 

[xp1632]

• Let's see why it loads wrong destination metadata json.loads('[{"colorChannel":"bgr","channelOrder":"channelLast","isMiniBatched":[true,false],"intensityRange":"0-255","device":"cpu"}]')

[xp1632]

Reason:

• After checking

• 

• We found the wrong destination metadata is from wrong defaultValue of Node edge_preserving_smoothing's input image handle

• 

• After update it to correct form, problem solved