Performance Evaluation

In this last part of the mid-term project you will document speed and number of matched keypoints for all detector / descriptor combinations with a focus on the preceding vehicle over a sequence of ten images.

TASK MP.7

Your seventh task is to count the number of keypoints on the preceding vehicle for all 10 images and take note of the distribution of their neighborhood size. Do this for all the detectors you have implemented.

TASK MP.8

Your eighth task is to count the number of matched keypoints for all 10 images using all possible combinations of detectors and descriptors. In the matching step, use the BF approach with the descriptor distance ratio set to 0.8.

TASK MP.9

Your ninth task is to log the time it takes for keypoint detection and descriptor extraction. The results must be entered into a spreadsheet and based on this information you will then suggest the TOP3 detector / descriptor combinations as the best choice for our purpose of detecting keypoints on vehicles. Finally, in a short text, please justify your recommendation based on your observations and on the data you collected.