In my opinion, to detect if the scaffold is from a particular organelle say mitochondria or chloroplast, you should have either sequenced from the individual organs or organelles and assembled it. At this stage, using any assembler (either Kmer or overlapping depending on your read length) should be good enough to assemble the organelle genome.

But, If the organs are pooled and the data is assembled, then the best way would be comparing with existing organ reference genome or have a set of transcripts say all the mitochondrial genes (or proteins) and see if all of them are present in the same scaffold you assembled.

There's a quite widely used tool named MITObim.

I know it's an old question, but it keeps popping up in the search, so I figured I'll update it.