Giter Site home page Giter Site logo

visual-sensor-fusion's Introduction

Visual-Fusion

  • LiDAR Fusion with Vision
  • Data taken from KITTI Dataset
  • Download Yolov4 model weights from here

Low-Level Fusion

Yolo Detections

Visualizing 3D LiDAR points in Open3D

3D Lidar Points Projected on the image plane

LiDAR points Fused with YOLO detections

  • LiDAR points are projected on the image using camera instrinsic and extrinsic matrix
  • The points that lie within the detected 2D Bounding Box by YOLO are stored and rest are ignored
  • There are some outliers inside bboxes that do not belong to that category, to reject these outliers there are several ways.
  • One way is to shrink the bounding box size so that the points that absolutely belong to the desired objects are only considered.
  • Another way is to use the Sigma Rule, i.e include the points that are within 1 sigma or 2 sigma away from gaussian mean, based on the distance of points

Mid-Level Fusion

Yolo Detections

LiDAR Points projected on Image

3D Bounding Boxes From LiDAR

3D BBox converted to 2D BBox

LiDAR 2D BBox Fused with YOLO 2D BBox using Intersection Over Union

  • 2D Bboxes from LiDAR are associated with YOLO 2D Bboxes using Hungarian Algorithm
  • Green Bounding Boxes are detected by YOlO whereas Blue Bounding Boxes are calculated using LiDAR points
  • YOLO missed 1 vehicle, whereas 2 vehicles are missed by LiDAR, one of which is half out of frame, at the bottom right side

File Structure

.
├── Code
|  ├── main.py
|  ├── Fusion.py
|  ├── Lidar2Camera.py
|  ├── YoloDetector.py
|  ├── Utils.py
|  ├── FusionUtils.py
|  ├── LidarUtils.py
|  ├── YoloUtils.py
├── Data
   ├── calibs
   |  ├── 000031.txt
   |  ├── 000035.txt
   |  ├── ...
   ├── images
   |  ├── 000031.png
   |  ├── 000035.png 
   |  ├── ...
   ├── labels
   |  ├── 000031.txt
   |  ├── 000035.txt 
   |  ├── ...
   ├── models
      ├── yolov4
      |  ├── yolov4.cfg
      |  ├── coco.names
   ├── output
      ├── images
      ├── videos
   ├── points
   |  ├── 000031.pcd
   |  ├── 000035.pcd 
   |  ├── ...

TODO

  • Add Run Instructions
  • Add Dependencies
  • Add References
  • High-Level Fusion

visual-sensor-fusion's People

Contributors

mjoshi07 avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar

visual-sensor-fusion's Issues

Error in reading Yolo Model

Small indexing error in YoloDetector.py line 21 that results in an error

Issue:
def get_layers(self):
ln = self.model.getLayerNames()
return [ln[i**[0]** - 1] for i in self.model.getUnconnectedOutLayers()]

Fix:
def get_layers(self):
ln = self.model.getLayerNames()
return [ln[i - 1] for i in self.model.getUnconnectedOutLayers()]

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.