1
1

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?

More than 3 years have passed since last update.

【3D OD Stereo】Pseudo-LiDAR : StereoDepthでPoint Cloudの生成

Posted at

Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving

![image.png](https://qiita-image-store.s3.ap-northeast-1.amazonaws.com/0/482094/674774f8-4240-f3bd-24c8-444540ae3d35.png)

Depth MapからPseudo LiDARはただの座標変換で計算できるので、今回はどうやってPseudo LiDARとImageで3D Object Detectionが出来るのかどうか見ていきたい。

新規性

3D Object Detection

LiDARで使われていたModelなら何でも良いらい この論文で2種類のモデルを試した。 ・Frustum-PointNet ・AVOD

Frustum-PointNet

![image.png](https://qiita-image-store.s3.ap-northeast-1.amazonaws.com/0/482094/3746d818-d157-333b-718f-04359cff1f93.png) 2D Object DetectionされたROIのPoint Cloudに対してPointNetを行う事で3次元BBを推定する方法。

AVOD

![image.png](https://qiita-image-store.s3.ap-northeast-1.amazonaws.com/0/482094/dbb938aa-f28d-5fe6-fe25-6ff3b929d2da.png) Imageと3D Anchor GridとBEVのpoint cloudを入力にして、物体を検出する2Stage Detector 詳細は別の記事書きます。

結果

![image.png](https://qiita-image-store.s3.ap-northeast-1.amazonaws.com/0/482094/57c5c6bb-325f-d1a0-b748-64018755b953.png) もちろんLiDARを使ったNetworkより精度はかなり落ちるが、改善の余地はある。
  1. Instance Segmentationを行いPoint CloudのFilteringをより正確に
  2. Stereo Depthの精度を高める

StereoDepthがLiDARの精度を超える事は考えにくいが、点密度が高い事が貢献してくれればと思う。

結論

・Stereo Cameraでもそこそこの精度で3次元物体検出は出来る。

参考文献

Pseudo-LiDAR from Visual Depth Estimation:Bridging the Gap in 3D Object Detection for Autonomous Driving https://arxiv.org/pdf/1812.07179.pdf
1
1
0

Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up
1
1

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?