OpenVL is a developer-friendly abstraction designed for computer vision across various processing units like APU, GPU, and CPU, addressing the gap between mainstream developers and complex computer vision technologies. It features a high-level API focused on problem descriptions rather than specific algorithms, allowing higher accessibility for non-experts while providing the flexibility and performance necessary for computer vision applications. The framework aims to promote widespread adoption by simplifying the use of existing computer vision methods and fostering innovative applications in the field.