Mild cognitive impairment (MCI) is a neurodegenerative disorder marked by cognitive decline that exceeds what is typically anticipated for an individual's age and educational background. Around half of those diagnosed with MCI transition to a diagnosis of dementia within a five-year period. Early detection of MCI is an area of study that is gaining traction due to its profound effect in improving the lives of people suffering from it and in helping to potentially delay the onset of dementia. This dissertation introduces a set of computational techniques using multiple cameras to detect behavior markers of group movement and social interaction, which can help distinguish various levels of cognitive functioning within a group environment. This is achieved by modeling and estimating the movements and social interactions of individuals with MCI within an indoor environment, while taking measures to uphold the privacy of the occupants.