Analysis‐aware microscopy video compression

This article introduces an analysis‐aware microscopy video compression method designed for microscopy videos that are consumed by analysis algorithms rather than by the human visual system. We define the quality of a microscopy video based on the level of preservation of analysis results. We evaluated our method with a bead tracking analysis program. For the same error level in the analysis result, our method can achieve 1,000× compression on certain test microscopy videos. Compared with a previous technique that yields exactly the exact same results by analysis algorithms, our method gives more flexibility for a user to control the quality. A modification to the new method also provides faster compression speed.