This is the code that produces that image. A mouse motion listener has been added to color the area under the pointer to dark green. Then running the shape algorithm based on colors near white (be patient - it takes a while), will produce a series of areas used to render this over the top.

All but the flood filling done in Java code not shown. Original image, cropped and reduced to a binary (2 color) image, then the oceans flood filled.

On the other hand, it has little effect on the possibility that some of the remaining coastlines depict Australia.A 'little' bit of number crunching can define areas or shapes of particular colors in an image.

The two types of evidence, toponymic and spatial, both of which are consistent with eastern Sumatra, suggest that previous proposals for the identification of this section of coastline cannot be maintained. The correspondences determined by place-name evidence are then used to conduct comparisons of spatial data, namely the latitudinal distances between the features of the Java la Grande coastline and the landmarks of eastern Sumatra. Many of the place-names are understood as native names or descriptions of Sumatran landmarks, without having to assume much textual error or to supply elaborate explanations. In this paper it is proposed that the east coast of Sumatra is the origin of a section of the western coastline of Java la Grande. It is otherwise thought to be made up of coastlines translated from elsewhere, notably Southeast Asia or South America, or it was invented for a cosmographical purpose. The earliest and most persistent theory holds that it is evidence of an early sixteenth century Portuguese discovery of Australia. Its identification has long been controversial. Abstract : In the mid sixteenth century, world maps and atlases produced at or near Dieppe on the Channel coast of France, a land mass called Java la Grande is depicted south of the Indonesian archipelago.

Java and sumatra on world map