GAC

本文以自主驾驶车辆为实际应用背景．提出了一种改进的ＲＲＴ（快速随机搜索树）路径规划算法．该路 径规划算法将非完整性约束条件与双向多步扩展ＲＲＴ搜索算法相结合，在提高搜索效率的同时保证了规划路径的可 行性．同时将路径点作为Ｂ样条基函数的控制点，用三次Ｂ样条函数来拟合控制点生成平滑可跟踪的路径．通过在平 面障碍物环境下实验，验证了该算法的有效性．(ＡｉｍｉｎｇａｔｔｈｅａｐｐｌｉｃａｔｉｏｎｏｆＡＬＶ（ａｕｔｏｍａｔｉｃｌａｎｄｖｅｈｉｃｌｅ），ａｎｉｍｐｒｏｖｅｄＲＲＴ（ＲａｐｉｄｌｙＲａｎｄｏｍ?ｅｘｐｌｏｒｉｎｇ Ｔｒｅｅｓ）ｐａｔｈｐｌａｎｎｉｎｇａｌｇｏｒｉｔｈｍ ｉｓｐｒｏｐｏｓｅｄ．Ｔｈｅａｌｇｏｒｉｔｈｍ ｃｏｍｂｉｎｅｓｔｈｅｎｏｈｏｌｏｎｏｍｉｃｃｏｎｓｔｒａｉｎｔｓｏｆｖｅｈｉｃｌｅｗｉｔｈｄｏｕｂｌｅｅｘｔｅｎｄ ＲＲＴｓ，ｗｈｉｃｈｎｏｔｏｎｌｙｉｍｐｒｏｖｅｓｔｈｅｅｆｆｉｃｉｅｎｃｙｏｆｓｅａｒｃｈｉｎｇｂｕｔａｌｓｏｇｕａｒａｎｔｅｅｓｔｈｅｆｅａｓｉｂｉｌｉｔｙｏｆｔｈｅｐａｔｈａｔｔｈｅｓａｍｅｔｉｍｅ．Ｔｈｅｐａｔｈ ｐｏｉｎｔｓａｒｅｔａｋｅｎａｓｒｅｆｅｒｅｎｃｅｐｏｉｎｔｓｆｏｒｔｈｅＢ?ｓｐｌｉｎｅｂａｓｉｃｆｕｎｃｔｉｏｎ，ａｐｐｒｏａｃｈａｎｄｃｒｅａｔｅａｎｅｗｓｍｏｏｔｈｒｏｕｔｅｗｈｉｃｈｓｕｉｔｓｆｏｒｔｈｅｖｅｈｉ? ｃｌｅｔｏｆｏｌｌｏｗ．Ｔｈｅｓｉｍｕｌａｔｉｏｎｏｕｔｃｏｍｅｓａｎｄｆｉｅｌｄｔｅｓｔｓｖｅｒｉｆｙｔｈｅｅｆｆｅｃｔｉｖｅｎｅｓｓｏｆｔｈｅｍｅｔｈｏｄｐｒｏｐｏｓｅｄ．)

多种刚性和非刚性配准算法，带有举例说明，非常好用(Multiple Rigid and Non-Rigid Registration Algorithms)

用IDl编写的，用IDL图像处理。已调试好的程序。(Prepared using IDL with IDL image processing. Debugging process has been good.)

采用拉普拉斯金字塔的图像融合程序，完成图像融合功能，需采用高于matlab2009的matlab(the image fusion source code using lap pyramid, you have to use the matlab version later than 2009.)

IMM算法，用于机动目标的精确跟踪和定位，该模型集考虑了三种典型子模型:CV,CT,CA. (interacting multiple model (IMM) code for target tracking )

实现简单图像的分块处理，可以对大图像或者目标复杂图像分成小块单独处理(To achieve simple image of the block processing, you can large images or targets complex images into small pieces, separate processing)

噪声估计法，先定位噪声点，再实行去噪处理(Noise estimation method, first locate the noise points, then implement denoising.)

主要用来对矩阵进行奇异值分解，与MATLAB有同样的效果(It is mainly used to perform singular value decomposition on matrices and has the same effect as MATLAB.)

将RGB格式的图片进行mosaic处理得到RawData格式的图片，然后再恢复成RGB格式彩色图片,laroche-CFA插值算法(turn an RGB picture to RawData picture，then recover the RawData picture to an RGB picture.it is called laroche-CFA.)

把PCNN和非下采样结合，不仅降低了所用时间，而且提高了效率。(Under the PCNN and the sampling combination, not only reduced the time and improve the efficiency.)