自适应分层粒子群稠密视差图估计

Dense Disparity Map Estimation Using PSO Algorithm with Adaptive Hierarchical Images

分析了遗传算法的缺陷,提出了自适应分层粒子群(PSO)立体匹配算法计算稠密视差图。首先采用SIFT(scale invariant feature transform)特征检测和匹配算法准确地确定视差范围;其次根据图像和视差范围的大小分层,建立由粗及细的自适应分层图像金字塔结构,加快搜索速度、减少错误匹配;然后在优化函数中引入能根据匹配窗口大小自动变化的因子来调整灰度项和平滑项数据的权重,并用改进的带变异算子的整数形式的PSO进行优化,避免了遗传算法搜索的盲目性以及容易陷入局部最优的缺陷,更快、更好地找到最优解。最后合成图像以及真实图像的实验结果表明该方法精度较高,速度较快。

An approach to addressing the stereo correspondence problem is presented using particle swarm optimization algorithm with adaptive hierarchy to obtain a dense disparity map. Firstly, the image features are precisely extracted by using SIFT feature detection, and accurately matched by using SIFT matching algorithm, so the disparity range is rightly and easily calculated from matching features. Secondly, according to restriction of the image size and the disparity range, the coarse to fine adaptive hierarchical image pyramid is built to search fast and reduce wrong matching. Thirdly, a regulation parameter varying with matching window is used to give different power for grayness and smoothness data in optimization function while the matching window is different in dissimilar supporting areas, and improved particle swarm optimization algorithm with variation operation for integer is used to find the fittest solution from a set of potential disparity maps avoiding Genetic algorithm＇ s blind searching and easy getting in local best solutions. Finally, experimental results on synthetic and real images show that the proposed approach performs dense disparity estimation accurately and quickly.