Search for tag: "artificial intelligence"

We start with an introduction to the set of videos, summarizing the six prototype computer vision systems.

We review coordinate geometry reference frame transformations and how to represent these.

We introduce a simple polycurve model that will be used to represent flat shapes and later extended for 3D shapes.

With the model to data correspondences, we can estimate the 2D transformation that maps the model onto the data.

Because the matching and pose estimation phases can produce spurious matches and poses, we introduce a verification step that eliminates bad hypotheses.

We next turn the boundary edge points into a more compact set of straight line segments.

First we find the edge points that surround the part, by analyzing a binary image.

Then we match those straight line segments to a rigid geometric model using the Interpretation Tree model matching algorithm.

When the background is changing due to lighting or leave motion, etc, simple subtraction is inadequate.We introduce a non-parametric statistical approach to modeling the background and then detect…

We introduce a simple flat rigid part recognition system that is based on matching part edges to a geometric model. This will also require estimating the reference frame transformation that maps the…

The Kalman filter is applied to the falling ball tracking problem.

We extend the ball tracking example to use Condensation tracking, which allows tracking of the ball through bounces and stopping.

When a target can change state, tracking can lose the target. If you have a model of different possible successor states, then the Condensation tracking algorithm can explore these…

We demonstrate the adaptive change detection algorithm on an indoor scene where the lighting is slowly changing, and where there are compression artifacts in the video.

We introduce the Kalman Filter, which is an important algorithm for estimation of variables given data that arrives periodically. Here it is used to track the ball, predict its position and refine…

We present several heuristics from removing small detections that arise because of noise, compression artifacts and camera jitter.