Since radiographic testing weld image possesses several shortcomings, such as bad contrast ratio, narrow range of grayscale and fuzzy image, common enhancement method cannot improve contrast ratio and at the same time preserve edges well. Fuzzy enhancement does not modify pixel graylevel in the fuzzy feature space just like histogram and can obtain high-definition image output. In this paper, on the basis of analyzing disadvantages of traditional fuzzy enhancement algorithm a generalized fuzzy enhancement algorithm is put forward. This algorithm can map the image to the generalized fuzzy space through involving the concept of generalized fuzzy set. At the same time, subsection sine function is chosen as fuzzy membership. Using the characteristics of generalized fuzzy transition with big range, image can obtain satisfactory enhancement effect through processing the radiographic testing image with generalized fuzzy enhancement algorithm. And this paper also introduces the method of fuzzy entropy to evaluate the enhancement effect and analyze the characteristics of sine membership function and generalized fuzzy enhancement operator.