Loading repository data…
abderrahmanefrt / repository
A computer vision project focused on 3D scene reconstruction using a stereo-vision system. This implementation involves camera calibration, image acquisition via translational motion, SIFT feature detection and matching, and 3D point cloud generation. Developed as part of the Master Informatique Visuelle curriculum at USTHB.
A transparent discovery signal based on current public GitHub metadata.
This score does not audit code, security, maintainers, documentation quality, or suitability. Verify the repository and its current documentation before adoption.
This project implements a complete pipeline for 3D reconstruction from stereo images (left/right). It estimates depth (Z) and generates a 3D point cloud in millimeters.
The system is based on stereo vision:
[ Z = \frac{f \cdot B}{d} ]
Where:
Z = depthf = focal length (pixels)B = baseline (mm)d = disparityLoad intrinsic matrix K
Convert portrait → landscape
Parameters:
fx, fy)cx, cy)cv2.findEssentialMat(...)
cv2.recoverPose(...)
Recovers:
Rtcv2.stereoRectifyUncalibrated(...)
Selected from shared topics, language and repository description—not editorial ratings.
AbhinandanVellanki /
Calibrating a camera for computer vision. Callibration done by calculating projection matrix, transformation matrix, essential matrix and camera calibration matrix
29/100 healthK_rect = H @ K
disp = pts_l[:,0] - pts_r[:,0]
cv2.triangulatePoints(P1, P2, ...)
Using:
P1 = K[I | 0]
P2 = K[R | t]
Output is in homogeneous coordinates → converted to 3D
scale = BASELINE / ||t||
pts3d *= scale
Remove:
Apply outlier filtering (3σ rule)
2D projections (X-Z, Z-Y)
3D scatter plot
Export:
resultat_3d.pngnuage_points.plyimage_left_undist4.jpgimage_right_undist4.jpgcamera_K.npycamera_dist.npy (optional)python main.py
3D point cloud
Generated files:
points_3d.npypoints_3d.txtnuage_points.plyresultat_3d.pngAccuracy depends on:
Translation vector t is normalized → scaling with baseline is required
stereoCalibrate)This project was developed as part of studies in computer vision and stereo vision.