| Twentieth century, HSE management theory has a good performance in the petrochemical industry, which is the preferred management system to chemical plant construction. With the continuous improvement of enterprise management level and the gradual accelerations of the information construction, enterprises further expand the application range of the HSE Management System. The HSE surveillance platform was proposed in order to implement the HSE management theory in primary production workshop. The functional requirements of platform is that, to rely on monitoring instruments to obtain the required data information, to analysis data by computer and to operate the controller or executing agency according to the computer’s conclusions, finally achieve the intelligent monitoring of the whole plant without human intervention. To meet requirements, the intelligent HSE monitoring platform consists of multiple subsystems. Image information as an important way for human access to outside information must be part of this platform.According to the platform’s requirement, this paper completed Production Safety Video Surveillance system to monitoring whether staff wears the safety helmet in the chemical plant. In order to achieve the Video Surveillance, the system should be video image recognition. Thus, first, study traditional video monitoring systems’structure, and choose devices for the intelligent monitoring system. Next, further study of the image processing, machine learning and pattern recognition theory, and design a helmet intelligent recognition scheme to achieve intelligent recognition in the production safety video surveillance system. In the scheme, using the interframe difference video object segmentation method to judge dynamic target; then using HOG characteristics to image and SVM classifier to determine whether the dynamic target images are pedestrians; With Adaboost classifier and Haar-like feature to detect whether the pedestrian is wearing a helmet. Finally, achieve the monitoring function for production safety video surveillance for the safety helmet.The Production Safety video surveillance system was test in experimental environment. The result shows that the scheme meets requirements and achieves the desired accuracy and reliability. This system applies the video recognition technology in HSE monitoring platform and achieves good effect, which makes the HSE management more intelligent. |
