The invention provides a recognition algorithm evaluation method and system based on infrared image big data. The method and system are used for solving the problem that in the prior art, an infraredimage recognition algorithm evaluation method is not high in applicability. The recognition algorithm evaluation method based on infrared image big data comprises the following steps: S1, establishingan infrared image test database; s2, calling a to-be-evaluated identification algorithm, selecting a to-be-tested image sequence from the infrared image test database according to the algorithm, andspecifying an identification target; s3, marking each frame of image of each sequence in the selected to-be-tested image sequence; s4, performing target identification on the selected to-be-tested image sequence by adopting the called identification algorithm; and S5, comparing the identification result in the step S4 with the marking result in the step S3, evaluating the identification algorithm,and outputting a result. According to the invention, a unified performance evaluation platform is provided for an identification algorithm in far and near scenes, automatic evaluation is achieved, and the evaluation efficiency and precision are improved.