Question about the Depth Morph Step.

[ruili3]

Hi,

Thank again for your brilliant work. I am a little confused when I read the Depth Morph part of the paper. In the implementation, what I understand is that you find the pixels $x$ around depth boundaries and then compute their corresponding pixel positions $x^{}$ according to Eq.(7), then the depth of $x$ are replaced by the depth of $x^{}$ to complete the morph.

I wonder whether I was right in understanding the idea of your paper. Thanks!

[TWJianNuo]

Hi!

Yes, the $x$ will be replaced by $x^{}$. To be more specific, the computation of $x^{}$ finally follow equation (9) to compromise among multiple depth-semantic edge pairs.

[ruili3]

Thank you a lot! I have one more question, then how to locate the corresponding multiple depth-semantic pairs when given a point at $x$?

TWJianNuo notifications@github.com 于2020年7月28日周二 下午8:46写道：

Hi!

Yes, the $x$ will be replaced by $x^{}$. To be more specific, the computation of $x^{}$ finally follow equation (9) to compromise among multiple depth-semantic edge pairs.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/TWJianNuo/EdgeDepth-Release/issues/5#issuecomment-665018348, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL6467WOLUVKTDCTOP254DR53CEDANCNFSM4PKHXLQQ .

[TWJianNuo]

For one semantic edge pixel, it will be paired with a nearest depth edge pixel. If no depth edge pixel present near it, then we skip. After matching, we suppress surrounding semantic edge pixel for efficiency consideration. Morphing computation is based on those computed pairs.

For each pixel, theoretically it will be affected by all pairs presenting in the image. But for ecfficiancy, we only search within a local range. The pairs beyond that range numerically has neglectable influence(e.g. 1e-3 change in pixel location).

[ruili3]

That's a very detailed explanation, thank you a lot!

TWJianNuo notifications@github.com 于2020年7月28日周二 下午10:24写道：

For one semantic edge pixel, it will be paired with a nearest depth edge pixel. If no depth edge pixel present near it, then we skip. After matching, we suppress surrounding semantic edge pixel for efficiency consideration. Morphing computation is based on those computed pairs.

For each pixel, theoretically it will be affected by all pairs presenting in the image. But for ecfficiancy, we only search within a local range. The pairs beyond that range numerically has neglectable influence(e.g. 1e-3 change in pixel location).

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/TWJianNuo/EdgeDepth-Release/issues/5#issuecomment-665071724, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL6463LDTAMGEI2OIPUHNTR53NS7ANCNFSM4PKHXLQQ .