Abstract: In the process of vision measurement,the limited data obtained by engineering prototype is applied to grey data mining. The grey data mining method and curve fitting are combined to realize the advantages of the two methods and then to find out the relationship between the position error and the measurement distance. First,a grey forecasting model,which can be applied to data mining,is built to enlarge the sample size of the limited data. Second,the grey correlation analysis is used to deal with the correlations of position measurement error. Finally,with the results of the grey forecasting model,a polynomial fitting of multivariate and multi-order is done. Experimental results show the data mining precision has reached first degree and the precision of curve fitting between the position error and the measurement distance is superior to 99%.