We have developed several systems to do image classification based on the scenic beauty quality of the images. For evaluation, we tested these systems under all kinds of conditions. Here are the results:

• Scenic Beauty Estimation: A collection of results of various classification algorithms on the Pre-Phase 01 data.

We are currently developing automatic image segmentation software. Our first step towards this goal involved a series of cheating experiments using manual segmentations. Subsequent research is focusing on the development of automatic classification techniques. Here is a summary of our experiments:

• Manual Segmentation: An overview of the experiments carried out on some specific regions segmented out from the Pre-Phase 01 images. The data set used was set 1. The segmentation was done manually with the aid of our labeling tool.
• Automatic Segmentation: A performance summary of our initial attempts at doing a block-level classification.

In the process of our research on image classification and segmentation, we did some optimization work on the edge and line detectors involved in our algorithms. Details are available here.