The application scope of image is gradually expanding，and related technologies are developing rapidly. The main support of its development is large-scale image data collection. The existing public resources are mainly image target data sets，while the background data sets are few，and the internal data volume is small，which restricts the development and application of image related technologies. Therefore，a Python expansion method of image background based on the characteristics of multi-stability and gray difference entropy is proposed. Based on the gray entropy difference between the image background and the target，the gray entropy threshold is selected to segment the original image；obtain the image background area；use the KNN algorithm to deeply mine the background data；apply the multi-stability characteristics to construct chaotic sequences；encrypt and process the image background data； develop a parallel expansion program for image background data through Python； and implement the established program to achieve the expansion of image background data. Experimental data show that the proposed method achieves better image background region segmentation effect， and the maximum image background data expansion is 1 800 MB， which fully confirms that the proposed method has better application performance.