"Factors like more crime being reported, more things being considered crime" - is there any reason to suppose either of these apply? How much of that rise do you think they could possibly account for, in the most extreme case?
Maybe I'm being ludicrous, but I think these factors could possibly have a huge effect on the statistics.
It would help clear things up if we could see a breakdown of crime by type. For example the need to report burglary for insurance purposes offers an economic incentive to do so, which may have a significant impact on the number of reported burglaries. The way drug law enforcement changed over the the century is another example. Increased phone ownership and mobility might mean that more incidents are reported to the police, Etc.
Another point to add is that the graph isn't per capita.