The focus of this DFG-funded projecton is on the hybrid estimation of non-functional properties—specifically, time, power, and energy—of software within embedded computer architectures used for image and video processing. The accurate determination of these properties, including runtime, electrical energy, and electrical power, is crucial given the limited computing resources available in embedded systems, as the algorithms and systems become more power-hungry. To optimize software performance during development, it is essential to evaluate its temporal behavior, as well as its energy and power consumption, particularly when operating with batteries. This project proposes a novel approach by combining two established estimation methodologies: a hardware-oriented approach based on processor energy models, and an application-specific analytical prediction method. By integrating these two methods, we aim to achieve precise estimations with minimal computational effort, ensuring efficient and effective evaluation of software non-functional properties. Our work leverages the RISC-V architecture to advance this research, setting a new standard for hybrid estimation in the field.