Localization algorithm
2016-08-23
1 0 0
no vote
Other
Earn points
This paper proposes a pedestrian detection method (HOG) based on the histogram of optical flow and direction gradient used by moving vehicles. The camera's egomotion is compensated after the same optical flow represented by the relative motion segmentation region extracted from the moving object. In order to obtain the optical flow field, two consecutive images are divided into 14 × 14 pixels, and then each cell tracks to find the corresponding cell in the next frame of the current frame. Using at least three corresponding cells, the affine transformation is performed according to each corresponding cell in the continuous image, so that the light flow is consistent with the extraction. The moving object in the region will be detected as the converted object, which is different from the previous registration background. The human habitation area of the candidate was obtained by morphological techniques. In order to recognize the object, pig feature extraction in the candidate region and classification, using linear support vector machine (SVM). Pig function vectors are used as input for linear svmto classification into pedestrian / nonpedestrian pairs for a given input. Proposed method was tested in
matlab
算法
定位
Related Source Codes
GMSK Linear Receiver
0
0
no vote
NSGA-II algorithm
0
0
no vote
NSGA-III multi-objective optimization algorithm
0
0
no vote
Compressed sensing example
0
0
no vote
CFAR detector example
0
0
no vote
No comment