This research outlines the development of a Bionic hands designed to emulate the way we use our fingers in a simplified and economical manner, through the use of computer vision as a means of capturing data. The development of the system utilizes real-time computer vision technology, allowing for hands-free recognition of hand gestures by using a camera to detect finger movements, with the data being analyzed by an image processing algorithm and provided to a microcontroller, thus controlling the movements of the servo motors which are located in each of the individual fingers. The design of the Bionic includes the construction via a 3D printer, which will result in a Bionic that will be light and can be custom made for each user. Each finger is actuated by means of a single-DOF linkage and is driven by a common multi-channel PWM (pulse width modulation) driver. The prototype exhibits excellent appearance and functionality (including finger movement accuracy) while also having fast response time and providing no need for complex calibration of sensors. Future improvements will include hybrid EMG signal integration, tactile feedback to the user, and adaptive learning so the user can better interact with their environment

This paper discusses the creation of a Bionic hand that uses vision to improve natural hand movement with less complexity, more cost effectively, and with higher accuracy of control. The system uses computer vision, an ESP32 microcontroller, and a PCA9685 servo driver for controlling the hand via servos connected through a single degree of freedom linkage to perform the grip/human-like movements of the fingers. The Bionic uses gesture detection through vision instead of using EMG- or contact-based sensing methods to provide smooth and precise motion while simplifying the control system. Because this Bionic is 3D printed, it has a lightweight, compact and configurable design, making it an appropriate solution for being a low-cost Bionic device. Future enhancements to system performance will be focused on integrating EMG and vision controlling in hybrid control techniques to provide increased accuracy and responsiveness of fingers. The presence of tactile feedback will provide increased control over gripping and handling items. Additionally, adaptive learning algorithms can provide a more individualized response from the Bionic device based on user needs. Currently, the system is limited by requiring external vision processing, and by incorporating onboard AI modules that will enable faster and fully independent operation, improvements can be made. The development of new Bionic hands will incorporate more advanced features, including multi-degree-of-freedom (MDF) finger mechanisms, improved actuator designs (i.e., electromagnetic- actuated) for providing actuation to the MDF fingers/tips through the manipulation of an external signal, and wireless communication that allows using the Bionic hand in a portable manner. Additionally, these added capabilities can be extended to use within a variety of applications, including robotic manipulators, rehabilitation devices, and human-machine interaction research. All enhancements will provide increased flexibility and usability during real-time application. Consequently, the Vision-Based Bionic Hand has shown that integrating computer vision, embedded systems, and 3D printed mechanical design can produce a more affordable and more effective replacement for existing Bionic hand control options. By reducing the complexity of systems, improving access to systems, and showing great promise for future intelligent Bionic devices that provide increased user comfort, safety, and functional capability, this solution is an innovative approach to improving the lives of amputees.

[1]M. Barandas et al., “AI-Enhanced Analysis to Investigate the Feasibility of EMG Signals for Bionic Hand Force Control Incorporating Anthropometric Measures,” BioMedInformatics, pp. 1–15, 2024. [2]L. Zhang et al., “A Bionic Hand System with Human- Like Grasping by Combining EMG and Visual Information,” IEEE Transactions on Neural Systems, vol. 32, no. 2, pp. 210–225, 2024. [3]S. Kumar et al., “Functional Evaluation of a Real-Time EMG Controlled Bionic Hand,” PMC Biomedical Engineering, vol. 18, no. 4, pp. 301–315, 2025. [4]M. Diab et al., “Development of a Low-Cost Bionic Hand Using Electromyography and Machine Learning,” arXiv preprint, pp. 1–12, 2024. [5]C. Cipriani, C. Antfolk, M. Controzzi, et al., “Online Myoelectric Control of a Dexterous Hand Bionic by Transradial Amputees,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 19, no. 3, pp. 260–270, 2011. [6]D. M. Barros and R. A. Lins, “Comparative Study of Mechanical Grippers with Emphasis on Underactuated Mechanisms for Bionics,” Journal of Mechanical Science and Technology, vol. 32, no. 10, pp. 4887–4896, 2018. [7]A. S. Soman, “Design and Fabrication of a Low-Cost Mechanical Bionic Hand,” International Journal of Research in Engineering and Technology, vol. 4, no. 1, pp. 27–31, 2015. D. M. Barros and R. A. Lins, “Design, Optimization, and Evaluation of a 3D-Printed Underactuated Bionic Hand,” Journal of Biomechanical Engineering, vol. 139, no. 12, p. 121004, 2017.

© 2025 International Journal of Engineering and Techniques (IJET).

Digital Object Identifier (DOI)

IJET assigns a unique DOI to every accepted article via Zenodo for persistent linking and citation. Your article’s DOI: https://doi.org/{{doi}}

Open DOI About Zenodo DOI

Close