Evolusi teknik denoising mesh: dari filter sederhana hingga pembelajaran mendalam 3D

Dalam artikel ikhtisar ini, Anda akan mempelajari apa itu denoising mesh, metode apa yang telah digunakan dan digunakan untuk menghilangkan noise pada mesh poligonal (dari filter klasik hingga jaringan neural grafik lanjutan), dan mendapatkan gambaran umum tentang pengembangan arah.

Mengapa Anda membutuhkan denoising mesh?

3D- . , ? Twin3d ( ) 3D-, . , , //. , .

3D- . reverse engineering CAD- , . CAD-, — , , , . , .

3D-, , mesh denoising. smoothing, «». 3D- , , . 3D-? , .

- … , (Laplacian smoothing, Taubin smoothing).

Menghaluskan Laplacian — Laplacian smoothing

2003 Bilateral mesh denoising – ( 2D ) . – , : , . 7 , (Bilateral normal filtering for mesh denoising), .

Bilateral Normal Filtering :

$n_i ^ k = \ Lambda (\ sum_ {f_j \ in N_i} A_jW_s (|| c_i-c_j ||) W_r (|| n_i-n_j ||) n_j),$

T_i — f_i , n_j — f_j , A_j — f_j , c_j — f_j ( ), $W (x) = exp (-x ^ 2 / (2 \ sigma ^ 2))$ — , $\ Lambda$ — .

, . W_s (x) — f_i f_j — , . W_r (x) , , . A_j

2015 Guided Mesh Normal Filtering, .

? , , . 3D- , .

mesh denoising, . , , Mesh Denoising via L0 minimization , . , – CAD- .

mesh denoising

2016 Mesh Denoising via Cascaded Normal Regression, . , noisy ground truth (GT) 3D- ( ).

(Synthetic) (Kinect v1, Kinect v2, Kinect Fusion). Synthetic noisy ( ). , . GT Artec Spider c , Microsoft Kinect.

Contoh dari dataset

– noisy ground truth GT . – . , . , . , .

Pipa Regresi Normal Bertingkat — Cascaded Normal Regression Pipeline

? -, . -, . , . , mesh denoising.

NormalNet: Learning based Guided Normal Filtering for Mesh Denoising. – , Guided Mesh Normal Filtering. (voxel – ), CNN. .

Pipeline NormalNet — NormalNet pipeline

NormalNet. CNN 4- , , . () — . ResNet, , . . , .

Arsitektur CNN di NormalNet — CNN NormalNet

, CNN , Guided Normal Filtering, . , Guided Normal Filtering.

, DNF-Net: a Deep Normal Filtering Network for Mesh Denoising . , end-to-end . ResNet , .

Arsitektur DNF-Net — DNF-Net

DNF-Net : . multi-scale feature embedding unit . . (. ) , ( ). F .

Unit Penyematan Fitur Multi-skala — Multi-scale Feature Embedding Unit

residual learning unit . KNN (K-Nearest Neighbors) k . k . “” .

Unit pembelajaran sisa — Residual learning unit

, , “” . , , .

, — , .. , “” .

Deep learning

, , . , - . 3D DL : .

, Mesh Denoising with Facet Graph Convolutions end-to-end pipeline , (, ) – , . — .

Pipa Konvolusi Grafik Aspek — Facet Graph Convolution pipeline

U-Net, , .

Arsitektur Jaringan Konvolusi Grafik — Graph Convolution Network

Grafik konvolusi

U-Net ( ), .

, 3D . - , -: , .

mesh denoising, , , , , ( ), .

Twin3d , mesh denoising . , ;)

All Articles