The clarity-evaluation-function of out-of-focus blurred image is a key factor for auto-focusing by digital image processing and the accuracy and efficiency of the function need to be improved. It is found that the blurred image has large amount of pixels with low gradient while the focused image has obviously more pixels with high gradient than the blurred image after deep study of the gradients of various images. Therefore, we can threshold gradients in order to throw low gradients and remain high gradients. And then the advantages of focused image are obviously showed. A kind of clarity-evaluationfunction of out-of-focus blurred image is firstly proposed in the paper. It is the sum of Laplacian operator with threshold. The mathematical model is established individually, and the experimental results and analysis are also given. The clarity-evaluation-function of out-of-focus blurred image given here has the characteristics of non-deflection, single peak value, high sensitivity and SNR.