Random sample consensus(RANSAC) algorithm is one of the most widely used for robust fundamental matrix
estimator. Considering low efficiency of RANSAC, a random sample consensus algorithm based on probability analysis
is proposed. The algorithm reduces the quantity of fist random sampling, and incorporates a better subset of model using
preemption scheme. The probability that the sample belongs to the better model is defined in order to be incorporated into
the sample subset. After several iterations, the sample subset contains only inliers. Experimental results of the fundamental
matrix computation on both simulated and real data show the superiority of the proposed algorithm in precision and efficiency.