In today's fast-changing competitive environment, companies and organizations need to develop their services, applications, and business processes continuously in order to benefit more from time and resources. Radio Frequency Identification (RFID), as one of the technological innovations, is an autoid technology that enables many new application domains in diverse scientific fields and provides important opportunities to build efficient industrial systems. It is capable of identifying, locating, tracking and monitoring objects, and offers companies and organizations important advantages to develop efficient and successful business applications. In RFID systems, tag identification is a critical process. However, tag collisions occur when multiple tags transmit their IDs to a reader simultaneously, and thus affect the stability of the RFID system. Therefore, a powerful tag anti-collision mechanism is required to accelerate the tag identification process while minimizing the collision effects. In this paper, we presented a performance analysis of some of the tree-based tag anti-collision protocols (binary tree, query tree, adaptive binary tree, and adaptive query tree) to evaluate the tag identification process for the RFID system in an IoT environment. In our experiments, we observed that adaptive tree-based tag anti-collision protocols have better performance than the binary tree protcol and the query tree protocol.