Optical flow method is frequently used for information estimation of multi moving objects. However, it is so difficult to get accurate solution by problem of aperture or discontinuities. In order to solve this problem, we proposed one of multi moving ...
Optical flow method is frequently used for information estimation of multi moving objects. However, it is so difficult to get accurate solution by problem of aperture or discontinuities. In order to solve this problem, we proposed one of multi moving object tracking algorithm based on the combinatorial Hough Transform (CHT) and voting accumulation, and considered operation time and error analysis.
In this paper, we extracted moving areas of moving objects from differential image using edge detection, filtering, region extension, and segmentation for each moving area. After finding real moving objects, Combinatorial Hough Transform (CHT) and voting accumulation is used to find the optimal constraint lines for optical flow estimation.
From optical flow vectors, we find the absolute vector for the direction and speed of the moving objects.
The simulation results show that the proposed method is very effective for extracting optical flow vectors and hence recognizing moving objects in the images. Moving objects are extracted to reduce the computation time for Combinatorial Hough Transform. The computational time is highly reduced by finding the moving objects using edge detection and segmentation. Optical flow constraints are solved using multi constraint and Combinatorial Hough Transform, which avoids the problems associated with the least-square based optical flow methods.