Research On The Detection Method Of Construction Tower Crane Based On DSOD Algorithm

As a pillar industry of the national economy,the construction industry has always played a pivotal role.However,construction accidents also occur frequently,with a high casualty rate.As the main type of casualty accidents,crane injuries,electric shock,object strikes,etc.are all caused by improper management of on-site construction equipment,with an accident rate of up to 25%.It can be seen that effective control of construction equipment can greatly reduce the occurrence of safety accidents.There are many types of construction equipment on the construction site,and tower cranes,as one of the main construction equipment on the construction site,are important potential factors for safety accidents such as falling from heights,object strikes,and crane injuries.Therefore,this article takes the construction tower crane as an example to study the effectiveness of the target detection algorithm.Firstly,this article introduces several common construction equipment in construction engineering and analyzes the difficulties in the detection of construction equipment.Then,starting from traditional target detection,the advantages and disadvantages of traditional detection methods and target detection based on deep learning were compared and analyzed.After learning that SSD series algorithms have the dual characteristics of high accuracy and fast detection speed,the DSOD algorithm in SSD series algorithms was selected and improved,while combining the types of construction equipment and the environment of the construction site,Taking the construction tower crane,one of the most important construction equipment in the construction site,as the detection object,a construction tower crane detection model based on DSOD-SE algorithm is proposed,and the feasibility of the improved algorithm is proved.In order to conduct integrated automatic detection of construction tower cranes,an automatic detection system for construction tower cranes was designed and developed based on the construction of an automatic detection model for construction tower cranes.The process of constructing image data sets,training and testing models,and analyzing detection results were integrated into a single system.The feasibility of the system in detecting construction tower cranes was demonstrated,providing a good platform for automatic detection of tower cranes in construction sites.From the perspective of equipment safety management in construction sites,this paper proposes a new method for detecting construction tower cranes,taking construction tower cranes as an example,providing a foundation for subsequent research on data collection of construction equipment and early warning of equipment hazards.At the same time,this article also applies computer vision technology to the field of construction engineering,enriching the knowledge system of vision technology,and providing a broader application range for information management in the field of construction engineering.The integrated development of computer technology and architectural engineering is a new trend and direction in the development of information technology such as deep learning and artificial intelligence.It also provides a theoretical basis and technical reference for research in other scientific fields in the field of architectural engineering.