它是基于一组包含异常数据的样本数据集，并计算数据的数学模型参数。这是第一次由Fischler和Bolles 1981提出。ransac是经常使用的计算机视觉。RANSAC假设为一组正确的数据，有一种方法，模型参数的计算。-It is based on a set of sample data sets containing abnormal data, and the mathematical model parameters are calculated. It was first proposed by Fischler and Bolles 1981. RANSAC is often used in computer vision. For example, in the field of stereo vision, the problem of computing the matching points and basic matrix is solved. The basic assumption of RANSAC algorithm is contained in the sample data (set), can be described by the model data, and data anomalies (outliers), deviation　the normal range, can not adapt to the mathematical model of the data. These abnormal data may be due to the error of the measurement, the assumption of error, the calculation error and so on. At the same time, the RANSAC is assumed to be a set of correct data, there is a way to calculate the parameters of the model.