Vision plays a key role as a synonym of scene-from-image reconstruction and understanding. In vision, spatial information typically dominates color information (Matsuyama and Hwang, 1990). This insight was ? and still is ? the foundation of geographic object-based image analysis (GEOBIA), proposed as a viable alternative to traditional pixel-based or local window-based 1D image analysis. In computer vision (CV), spatial concepts in the scene- and image-domain, such as local shape, texture, inter-object spatial topological and spatial non-topological relationships, have been investigated since the late 1970s (Nagao and Matsuyama, 1980). In GIScience, ?object-based image analysis? (OBIA) was tentatively introduced in 2006 ( Lang and Blaschke, 2006). In 2008, it was re-formulated as GEOBIA (Hay and Castilla, 2008) emphasizing a primary focus on Earth data-derived applications and the interdisciplinary novelty of geospatio-temporal reasoning to cope with massive Earth observation (EO) imagery, where to integrate achievements accomplished by the remote sensing (RS) and CV communities (Lang, 2008). By 2010, a plethora of published papers focused on the GEOBIA approach (Blaschke, 2010), following the drive of a new-borne OBIA commercial software eCognition (Benz et al., 2004), built upon a semi-automatic multi-scale image segmentation approach (Baatz and Schäpe, 2000). GIScience scholars proposed GEOBIA as a paradigm shift (Blaschke et al., 2014), capable of bridging the semantic information gap from big image data, such as long-expected very high resolution EO images through time, to information primitives in the real-world scene-domain to be handled by GISs (Lang and Blaschke, 2006). Early reflections on strengths and weaknesses of this ?new discipline? (Hay and Castilla, 2008) were followed by substantial criticisms. In detailed statements, Baraldi and Boschetti (2012) tried to pin down the promises of GEOBIA promoters against objective product and process quality indicators, including degree of automation, scalability, robustness, timeliness and costs (GEO, 2010). To recover from an ?unquestionable lack of productivity? of GEOBIA systems (Baraldi and Boschetti, 2012) [p. 33], this contribution highlights recent achievements, and open issues to be tackled by the GEOBIA community, possibly developing towards a GEOBIA 2.0 Manifesto. Flanked by recent trends, such as big data analytics challenged by the five v's of volume, velocity, variety, veracity plus value, we promote key aspects of GIScience stemming from multi-source EO image analysis, including: (1) EO image enhancement and interoperability/harmonization at the radiometric and semantic levels of analysis; (2) big raster data storage and exploitation, affected by ongoing lacks in tackling spatiotemporal information in vector format; (3) deep (multi-scale) distributed CV systems (artificial neural networks), capable of 2D topology-preserving (retinotopic) image feature mapping and provided with feedback loops, alternative to feedforward 1D image analysis; (4) hybrid (combined deductive/top-down and inductive/bottom-up) inference, (5) structured CV system of systems design, based on a convergence of spatial and color evidence, (6) consistency of CV with human visual perception; (7) semantic content-based image retrieval. Contributions to the ongoing GEOBIA 2.0 requirements specification should lead to a white paper (or similar) emerging from the GEOBIA 2018 Conference (and beyond) that positions the GEOBIA community within the interdisciplinary realm of big image data analytics.