The objective of the work is to detect shadows in images and identify the problem of labeling image regions, where each region corresponds to a group of super pixels. To predict the label of each region, the proposed work train a kernel Least-Squares Support Vector Machine(LSSVM) for separating shadow and non-shadow regions. The parameters of the kernel and the classifier are jointly learned to minimize the leave-one-out cross validation error. Experiments on two challenging shadow datasets, UCF and UIUC, show that the region classifier outperforms more complex methods.