In the past, I worked on many detection tasks using airborne images for the purpose of vision-based navigation of UAVs. I found out that when a vision system relies on detecting an ‘entire’ target, the navigation is not really robust. From time to time during target chasing, there are moments that the vision system sees only the part of the target due to tracking errors and disturbances, etc., and that absolutely happens at some point of vision-based landing because the target gets too large to fit into an onboard image. One approach to solve this issue with traditional machine-learning-based algorithms such as Viola-Jones and Dalal Triggs frameworks would be to add occluded images to positive images. Still, a naive implementation would lose a lot of information. For example, for the purpose of vision-based landing, detecting an occluded target is not enough, and we need to know where the center of the target is from occluded images. If we treat all the images that contain part of the target, we never know where the center of the target is. My approach is to use the combination of object proposals and particle filters. We have many algorithms to evaluate ‘objectness’ such as selective search, BING, etc. Each proposed object is compared with the target image using HOG features, normalized cross-correlation, and SIFT to obtain the score of each object. The measurements are fused via a particle filter framework.