SENSOR-INTEGRATED DIGITAL CANVASES ALLOWING ARTISTS TO MANIPULATE PAINTINGS THROUGH GESTURAL INPUTS

Abstract [English]

The digital canvases that are sensor-integrated is another new twist in the world of interactive visual art since artists can now dictate the digital paintings by natural gestures. The given paper gives a full architecture of the construction and introduction of such systems based on the integration of multimodal sensing technology like motion sensors, depth cameras, inertial measurement units (IMUs), and touch-sensitive interfaces. The proposed system architecture has a potent gesture processing unit and high quality rendering unit making access to art in real time easy. Gesture recognition and classification is a gradual process algorithm that is developed using machine learning and allows a correct analysis of hand movements and body actions. These gestures are dynamically assigned to the processes of art such as brush strokes, tweezing of textures, mixing of colors and geometrical changes. Moreover, the adaptive learning capabilities of the system also tailor the response to the user interaction patterns, making the system more expressive in the field of creativeness. An edited list of gestures and a calibrated hardware system is experimentally tested with significant improvements in workspace accuracy, response rate, and control compared to the traditional digital input applications. The results show how the system has the potential of rebranding the digital art production by offering more interactive, natural and responsive space to creative work across different artists.