Digital CR of head and neck overcomes the disadvantage of regular X-ray radiography, which can not reveal bone and soft tissue position deficiency in one exposing, and reduces the X-ray radiation dose. Meanwhile, various factors decline the imaging quality, and images must be enhanced in order to meet demands of doctors clinical diagnosis. The general enhancement algorithms don't consider bodys structure difference and density characteristics. A new adaptive CR enhancement algorithms was proposed in this article, and head and neck images were processed with this method and compared with linear unsharp masking method. The results proved that the details of CR image enhanced were abundant and enhanced CR had good visual effect. SNR was high, as well as detail variance/background variance (DV/BV) indicating that this algorithms is suitable for head and neck CR medical image.