论文部分内容阅读
介绍了一种基于图像识别的室外火灾智能报警系统,包括系统组成和工作流程。该系统运用差值图像的阈值化来识别可疑亮区,并运用膨胀算法对图像中的可疑亮区进行窗口分割。通过序列图像的采集以及可疑亮区的窗口变化,运用人工智能中的规则推理进行火灾的智能识别。
An outdoor fire alarm system based on image recognition is introduced, including system composition and workflow. The system uses the thresholding of the difference image to identify the suspicious light area and uses the dilation algorithm to segment the suspicious light area in the image. Through the acquisition of sequence images and the change of window of suspicious bright area, the intelligent recognition of fire is carried out by rule reasoning in artificial intelligence.