The fractional Foruier transform (FrFT) is introduced into the dual nonlinear optical correlation. The dual nonlinear fractional correlation can be realized by the nonlinear operations on the FrFT of the reference images and the target ones. The proposed scheme gives different weights to different shapes and textures by nonlinear parameters and fractional orders. Therefore the properties of target recognition can be controlled. The dual nonlinear fractional correlator can be implemented by an opto-electronical setup. The simulation results show that the most remarkable characteristic of this system is that it has the variable and adjustable discrimination capability to shape changes and texture distortions of the objects. In addition it can improve the performances of the correlation output peak and it has better rubustness to noise.