Given an approximately centered image of a spiral galaxy, we describe anentirely automated method that finds, centers, and sizes the galaxy and thenautomatically extracts structural information about the spiral arms. For eacharm segment found, we list the pixels in that segment and perform aleast-squares fit of a logarithmic spiral arc to the pixels in the segment. Thealgorithm takes about 1 minute per galaxy, and can easily be scaled usingparallelism. We have run it on all ~644,000 Sloan objects classified as"galaxy" and large enough to observe some structure. Our algorithm is stable inthe sense that the statistics across a large sample of galaxies vary smoothlybased on algorithmic parameters, although results for individual galaxies cansometimes vary in a non-smooth but easily understood manner. We find a verygood correlation between our quantitative description of spiral structure andthe qualitative description provided by humans via Galaxy Zoo. In addition, wefind that pitch angle often varies significantly segment-to-segment in a singlespiral galaxy, making it difficult to define "the" pitch angle for a singlegalaxy. Finally, we point out how complex arm structure (even of long arms) canlead to ambiguity in defining what an "arm" is, leading us to prefer the term"arm segments".
展开▼
