The image category strategy for an input cover image is considered as a part of the entire adaptive steganographic algorithm in the current image-based adaptive steganography. This results in two drawbacks; namely, the adaptive steganographic algorithms developed are inflexible to cope with the embedding techniques; the capacity provided cannot satisfy user's demand. In this study, we separate the category strategy and the embedding algorithm from the whole adaptive steganographic algorithm. We propose two adaptive steganographic algorithms to tackle drawbacks. The first algorithm is the flexibly adaptive multiple base steganographic algorithm. We introduce two statistical methods for the image category strategy including the mean of the pi...