Each day we open a door using physical keys or tokens like RFID or smart cards. While we all are used to these methods they have problems of security and usability. These tokens and keys can easily be stolen or taken by other persons which results in a security problem. The problem in usability is that users need a significant amount of time to take out their tokens to unlock and open the door. In this paper, we propose a new approach for door handle access control. We developed a prototype by attaching pressure sensors to the door handle that measure resistive and capacitive touch interactions with the door handle. We demonstrate the feasibility of identification with a door handle with a visual and classification analysis. The classification algorithms used are K-NN, SVM, Random Forests, AdaBoost and MLP achieving a maximum accuracy of 88% using the Random Forests.