On the basis of comprehensive analysis about limitations of Pal.King algorithm,an improved algorithm of fuzzy edge-detection is presented.Also proposed is a new method based on genetic algorithm to determine the optimal threshold of the membership grade.The simple membership function proposed in the paper simplifies the complex transformation calculation in Pal.King algorithm.In order to obtain the desired effect,the μc value is optimized according to the requirement.The edges of the image are extracted according to a rule combining the “Max” and “Min” operators.Simulation results show that with the proposed algorithm the quality of edge detection is largely improved and the computation of the algorithm is faster than that of Pal.King algorithm.