Correlation analysis-based image segmentation approach for automatic agriculture vehicle

It is important to segment image correctly to extract guidance information for automatic agriculture vehicle. If we can make the computer know where the crops are, we can extract the guidance line easily. Images were divided into some rec-tangle small windows, then a pair of 1-D arrays was constructed in each small windows. The correlation coefficients of every small window constructed the features to segment images. The results showed that correlation analysis is a potential approach for processing complex farmland for guidance system, and more correlation analysis methods must be researched.