Identification content detection method and system based on deep integrated network

The invention discloses an identification content detection method and system based on a deep integrated network. The method specifically comprises the following steps: acquiring a to-be-detected picture or video frame; inputting the to-be-detected picture or video frame into Q pre-trained sub-detection models, and performing reasoning in parallel to correspondingly obtain Q reasoning results; splicing the Q reasoning results to obtain a spliced reasoning result; carrying out filtering processing on the spliced reasoning result based on a preset threshold value to obtain a filtering result; and obtaining an identification content detection result based on the filtering result. The identification content detection method or system provided by the invention has relatively high detection efficiency and accuracy during identification content detection. 本发明公开了一种基于深度集成网络的标识内容检测方法及系统，所述方法具体包括以下步骤：获取待检测图片或视频帧；将所述待检测图片或视频帧输入预训练好的Q个子检测模型中，并行进行推理，对应获得Q个推理结果；将所述Q个推理结果进行拼接，获得拼接后的推理结果；基于预设阈值对所述拼接后的推理结果进行过滤处理，获得过滤结果；基于所述过滤结