GitHub - TouficKashmar/Motion-Estimation: Motion estimation is the process of determining motion vectors that describe the transformation from one 2D image to another; usually from adjacent frames in a video sequence. Usually, optical flow is used for this application. Optical flow is the pattern of apparent motion of image objects between two consecutive frames caused by the movement of object or camera. It is a 2D vector field where each vector is a displacement vector showing the movement of points from the first reference frame to a second.frame. The Lucas-Kanade (LK) and the Kanade-Lucas-Tomasi (KLT) algorithms are popular optical flow computation methods.

TouficKashmar / Motion-Estimation Public

Notifications You must be signed in to change notification settings
Fork 1
Star 0

Motion estimation is the process of determining motion vectors that describe the transformation from one 2D image to another; usually from adjacent frames in a video sequence. Usually, optical flow is used for this application. Optical flow is the pattern of apparent motion of image objects between two consecutive frames caused by the movement o…

0 stars 1 fork Branches Tags Activity

Star

Notifications

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 2 Commits
FAST_FREAK		FAST_FREAK
FAST_LUCID		FAST_LUCID
KLTOpticalFlow_Gray		KLTOpticalFlow_Gray
KLT_Gray_Arrow		KLT_Gray_Arrow
README.txt		README.txt

Repository files navigation

- Application: Motion Estimation
Motion is estimated by drawing the optical flow within two videos of moving cars
(https://www.youtube.com/watch?v=Y1jTEyb3wiI)(https://www.bogotobogo.com/python/OpenCV_Python/images/mean_shift_tracking/slow_traffic_small.mp4)

using three different approaches:
1. FAST keypoint detector algorithm with the pyramid implementation of the Lucas-Kanade (KLT)
algorithm (Optical flow. OpenCV. (n.d.). Retrieved November 7, 2021, from https://docs.opencv.org/3.4.15/d4/dee/tutorial_optical_flow.html).
2. FAST keypoint detector with the FREAK keypoint descriptor (descriptor matching).
3. FAST keypoint detector with the LUCID keypoint descriptor (descriptor matching).

- Steps to run KLTOpticalFlow_Gray:

1. Open the Code:Blocks IDE
2. Go to File --> Open KLTOpticalFlow_Gray.cbp file.
Note: "KLT_Gray_Arrow.cbp" is the same code but it outputs the optical flow using arrows.
Note: "slow_traffic_small.mp4" was also used to compare results. To use it, change the VideoCapture object name in the code.
3. Press the "build and run" button.
4. A console screen and a windows will appear showing the optical flow obtained from the pyramid implementation of the KLT algorithm.
5. Exit the program by pressing ESC.

- Steps to run FAST_FREAK:

1. Open the Code:Blocks IDE
2. Go to File --> Open FAST_FREAK.cbp file.
Note: "slow_traffic_small.mp4" was also used to compare results. To use it, change the VideoCapture object name in the code.
3. Press the "build and run" button.
4. A console screen and a windows will appear showing the optical flow obtained from using FAST + FREAK algorithms.
5. Exit the program by pressing ESC.

- Steps to run FAST_LUCID:

1. Open the Code:Blocks IDE
2. Go to File --> Open FAST_LUCID.cbp file.
Note: "slow_traffic_small.mp4" was also used to compare results. To use it, change the VideoCapture object name in the code.
3. Press the "build and run" button.
4. A console screen and a windows will appear showing the optical flow obtained from using FAST + LUCID algorithms.
5. Exit the program by pressing ESC.