docs: add historical charts to benchmarks readme

benchmarks/README.md

@@ -156,6 +156,19 @@ The worst memory usage is by cc3d with sparse mode off. Scipy and cc3d are appro
 | cc3d 1.14.0       |  339    |  3.36x     |
 | cc3d 1.14.0 sparse| 1107    | 10.96x     |
 
+# Historical Performance
+
+cc3d has been steadily improving over time. To celebrate the release of 2.0.0, we show plots of peak memory usage and megavoxels per second vs version.
+
+<p style="font-style: italics;" align="center">
+<img height=512 src="https://raw.githubusercontent.com/seung-lab/connected-components-3d/master/benchmarks/cc3d_26-way_connectomics_over_time.png" alt="Fig. 5: cc3d peak memory usage and speed in selected releases against a 512x512x512 connectomics dataset." /><br>
+Fig. 5: cc3d peak memory usage and speed in selected releases against a 512x512x512 connectomics dataset.
+</p>   
+
+<p style="font-style: italics;" align="center">
+<img height=512 src="https://raw.githubusercontent.com/seung-lab/connected-components-3d/master/benchmarks/cc3d_26-way_connectomics_over_time.png" alt="Fig. 6: cc3d and scipy peak memory usage and speed in selected releases against a 512x512x512 random binary dataset." /><br>
+Fig. 6: cc3d and scipy peak memory usage and speed in selected releases against a 512x512x512 random binary dataset.
+</p>