Hand posture-based remote control and human-machine interaction has been the subject of widespread research and development for decades, aiming to provide the ability to communicate with computer systems in a natural, human way. Lately, the OpenPose library has been used more and more in computer vision-based pose estimation applications (e.g. remotely controlling home appliances, robots, using hand postures and gestures). It processes camera images and builds a simplified skeleton of the estimated pose of the person in the camera frames. The resulting posture model consists of 2D keypoint coordinates and their confidence values, which are then used to identify the pose, typically with a deep learning method. However, deep learning methods usually require a large amount of data to train on and operate with a high computational complexity. In previous work, we have introduced the so-called Sequential Fuzzy Indexed Search Trees classifier that was designed for fast classification, even for high-dimensional data (such as pose models). It combines the generalization ability of fuzzy inference systems and the high speed of indexing table classifiers. In this paper, a hand posture identification framework is proposed that can be used for computer vision-based remote control or other human-machine interaction applications. It incorporates reliable posture model building with the OpenPose library and fast identification using the SFIST classifier. Experimental results are provided to show that the latter can provide a slightly worse (by ~1%) classification performance compared to two other state-of-the-art classifiers, but it does so much faster and at much lower memory cost.