What is the target detection algorithm?

The target detection algorithm is to learn a classifier through the training set, and then slide and scan the whole image with different scale windows in the test image; Classify each scan to determine whether the current window is the target to be detected. The core of detection algorithm is classification, and the core of classification is which features to use and which classifier to use.

Extended data:

Target detection algorithms can be divided into:

1. Background modeling methods, including time average model, mixed Gaussian model, dynamic texture background, PCA model and spatio-temporal joint distribution background model?

2. Point detection method, including Moravec detector and Harris detector? Affine invariant point detection, IFT

3. Image segmentation methods, including mean shift method, graph cutting method and active contour method.

4. Cluster analysis, including support vector machine, neural network and adaptive Boosting.

5. Motion vector field method, including the method based on motion vector field.