Abstract: In this paper we propose a method by dividing the original signal to multiple parts and assign new meaning to "1" and "0", to calculate Lempel-Ziv complexity at multi-scale without increasing the number of symbols. This new approach is presented for distinguishing textual images from pictorial images and we find that the complexity of pictorial images have more significant increase than that of textual images when the calculation scale becomes small. The test shows that multi-scale Lempel-Ziv complexity can be used as an image classification method or as a feature of image classifier.