 Infrared video from aerial platforms is increasingly being used to estimate wildlife densities. However, there are several sources of imperfect detection which need to be taken into consideration. These include errors caused by the viewer of the infrared video, as well as errors due to the thermal sensor itself. To address this issue, we suggest using double observer methods to assess imperfect detection, as well as combining infrared video with red-green-blue RGB video to reduce false positives. Additionally, it is important to apply stringent verification standards to ensure accurate detection and to collect data at lower altitudes over snow when needed. This article was authored by Zachary J. Deleel, Patrick G. McGovern, Brian G. Dilman, and others.