We study the use of visual content tracking applications in conjunction with blockchain solutions. The issue consists of multimedia technologies such as fingerprinting requiring the computing power of off-chain machines to operate. We present a workflow using a load-balancing architecture that allows for such operations to be computed off-chain yet benefitting from the level of integrity brought by blockchains. We thus enable the creation of blockchain-backed data storages where multimedia inputs can be identified by their semantic content. We also address the challenge of minimizing computing resources which occurs when on-chain processing is in context, ensuring such an architecture can be deployed and used on state-of-the-art environments. We illustrate our workflow on the Tezos and Ethereum platforms and provide two implementations, relying on an image filter detection use case (featuring the International Standard Content Code), and a museum IPR use case (featuring a robust video fingerprinting method), respectively. We thus demonstrate the mutually beneficial association of onchain and off-chain applications for multimedia content tracking as well as the blockchain-agnostic nature of the advanced solution.