Description: Based on novel Progressive Scale Expansion Network (PSENet),we train and light adjust to get a better result.As a segmentation-based method, PSENet is able to locate texts with arbitrary shapes. Besides, a progressive scale expansion algorithm is put forward, with which the closely adjacent text instances can be identified successfully. Specifically, each text instance with multiple predicted segmentation areas are assign. these segmentation areas are denoted as kernels and for one text instance, there are several corresponding kernels. Each of the kernels shares the similar shape with the original entire text instance, and they all locate at the same central point but differ in scales. To obtain the final detections, we adopt the progressive scale expansion algorithm，which is based on Breadth-First-Search (BFS) , starting from the kernels with minimal scales and expanding their areas by involving more pixels in larger kernels gradually, and then finishing until the largest kernels are explored.