Lumineyes Laser Surgery

feature extraction

CAMSHIFT and mean-shift algorithm in OpenCV OpenCV

used for tracking the CAMSHIFT algorithm. The advantage is that you can also use lower-cost webcams. Such a algorithm that work efficiently and quickly and track in real time, can not exploit the resources of the PC so that other applications can run alongside.
The mean shift algorithm is based on a robust parameter-free technique for growing Dichtheitsgradienten to the summit of the probability distribution to find.

We want a procedure for the distribution of colors found in a video scene.

For this reason, the mean shift algorithm has been modified to make it with dynamically changing color probability distributions to work, coming from the video image sequence. This modified algorithm is the Continuously Adaptive Mean Shift Algorithm (CAMSHIFT).

operation of the algorithms:
mean shift algorithm: first
Setting a search window size
second Set the start position of the search window
third Calculate the center position in the search window
4th Center the search window at this middle position
5th Repeat steps 3 and 4 until convergence (or up to a given threshold
)

The mean shift algorithm works with probability distributions in order to tracking colored objects in video sequences. The color image data must be probability distributions are presented. Color histograms are used to accomplish this.

CAMSHIFT algorithm: first
Set the start position of the second search window
Application of mean shift (see above), Save the zeroth moment
third Equating the search window size to a function of the zeroth moment, which was found in step 2
.
4th Repeat the 2nd and 3 Step up to an agreement ("zeroth moment".
distribution area below the search window window radius, height and width, in the
function used zeroth moment)

The CamShiftAlgorithmus track that X, Y and areas of skin color probability distribution. The Area is proportional to Z, the distance to the camera.