为提高图像数据压缩算法性能，提出一种基于可变比特率块编码无环图模型的图像无损压缩。首先，从零初始化索引开始，沿着从根到叶的路径遍历时，使用遍历过程中体素计数的运行和计算完整树遍历中特定节点之前的体素数量，获得连续表示的基于稀疏体素八叉树DAG体素索引模型；其次，在离线图像压缩格式和固定比特块编码之间的中间位置，引入一个二进制搜索来定位所含的块，使用比实际要求的请求错误阈值所需比特来描述图像数据块，实现更好的数据压缩性能；最后，通过仿真实验验证了所提算法在图像以及矢量图像数据压缩上的性能优势。

To improve the performance of image data compression algorithm, a lossless image compression based on variable bit rate block coding acyclic graph model is proposed. Firstly, starting from zero-initialization index and traversing along the path from root to leaf, the sparse voxel octree DAG voxel index model based on continuous representation is obtained by using the operation of voxel counting during traversal and the number of voxels before specific nodes in complete tree traversal. Secondly, in the middle position between the offline image compression format and the fixed bit block encoding, a binary search is introduced to locate the contained block, and the bit required to describe the image data block is used to achieve better data compression performance than the actual request error threshold. Finally, the performance advantages of the proposed algorithm in image and vector image compression are verified by simulation experiments.