Video monitoring can provide vital context awareness information from indoor intelligent environments where privacy is not a limitation. However, there is a need to develop linguistic summarization tools which are capable of summarizing in a layman language the information of interest within long video sequences. The key module which can enable the linguistic summarization of video monitoring is human activity/behaviour recognition. However, human behavior recognition is an important yet challenging task due to the behavior uncertainty, activity ambiguity, and uncertain factors such as position, orientation and speed, etc. In order to handle such high levels of uncertainties in activity analysis, we introduce a system based on Interval Type-2 Fuzzy Logic Systems (IT2FLSs) whose parameters are optimized by the Big Bang-Big Crunch (BB-BC) algorithm which allows for robust behaviour recognition using 3D machine vision techniques in intelligent environments. We present several experiments which were performed in real-world intelligent environments to fairly make comparisons with the state-of-the-art algorithms. The experimental results demonstrate that the proposed BB-BC paradigm is effective in tuning the parameters of the membership functions and the rule base of the IT2FLSs to improve the recognition accuracy. It will be shown through real-world experiments that the proposed IT2FLSs outperformed the Type-1 FLSs (TIFLSs) counterpart as well as other traditional non-fuzzy based systems. Based on the recognition results, higher-level applications will presented including video linguistic summarizations event searching and activity retrieval/playback.