Rectangular based binary image representation: Theory, applications, and dataset introduction

This paper presents the theory and applications of rectangular based binary image representation (RBIR). An automatic method is proposed to convert original binary image into RBIR. Some applications directly derived from RBIR are introduced. RBIR lossless compression is comparable with the run-length encoding (RLE) compression. Efficient algorithms are proposed for computing the dilation/erosion from the RBIR. Unfortunately, converting original image into the RBIR is not a fast process. Hence, a dataset is provided which researchers can use to report their results.