There are three 5T explosion-proof single beam cranes in the customer's factory, which are required to be able to solve personnel intrusion alarms in the hook area during operation and identify obstacles during crane travel, ensuring the safety of personnel and equipment during construction.

Technical difficulties

1. Low temperature and explosion-proof usage environment: The winter temperature in the area where the equipment is used is -20 ℃, and on-site treatment is required according to explosion-proof specifications.

2. Flow area monitoring: It is required to monitor the area below the hook for personnel intrusion during walking operations.

3. Top view angle AI recognition: The on-site camera is installed on the trolley, and the capture angle is perpendicular to the ground, making recognition difficult.

4. The radar signal should respond quickly while also addressing the issue of false alarms.

Technical solutions

1. Select suitable low-temperature and explosion-proof components for the customer's usage environment.

2. Traditional AI recognition combined with object detection technology completes the preprocessing task of video images in the monitoring area through steps such as background modeling, grayscale transformation, smooth filtering, image enhancement, and edge detection; According to the principle of binocular stereo imaging, the target disparity map is obtained by completing stereo matching of the target through the video image detection results of adjacent cameras; Determine whether the intruder has crossed the boundary to determine whether the incident should be alerted.

Traditional personnel intrusion cameras are inclined at a certain angle to capture the person's face for recognition, while the micro on-site camera is installed on a small car, and the capture angle is perpendicular to the ground, which brings great difficulty to capture. Micro technical engineers braved the severe cold weather of -20 ℃ and increased the amount of on-site material collection. Through algorithm optimization, they solved the problem of identifying personnel from top view angles and conducting continuous on-site testing, ultimately meeting customer requirements.

3. The anti-collision radar system monitors the lower part of the hook in real-time. When a dangerous target (such as a cargo rack or other stacked objects) appears, the radar system sends an alarm to the crane driver in advance, enabling the driver to respond in a timely manner. At the same time, the radar outputs a signal to the crane control system, which automatically stops or decelerates according to the situation.

By adjusting the angle of the on-site radar sensor and optimizing the algorithm, the radar signal can meet the requirements of fast and accurate recognition.

Product and Service Delivery

1. Micro specialty products have been exported to Russia and have long-term cooperation with customers such as nuclear power and CNOOC. The low-temperature and explosion-proof technical solutions are relatively mature, and product testing can be carried out before leaving the factory to ensure that the quality of the products meets the environmental requirements of customers.

2. Through on-site repeated testing and algorithm adjustments, the accurate identification and alarm of personnel intrusion into the dangerous area flowing below during the hook walking operation process according to customer requirements have been achieved on site.

3. Through on-site testing and adjustment of radar sensor transmission angle by micro special staff, it is provided to software engineers for continuous adjustment of software algorithms. The on-site program is updated no less than 10 times before and after. Finally, the obstacle detection direction forms a "back" shape, and the center and surrounding area of the back do not recognize obstacles. Only between the large and small openings can obstacles be recognized.

Use feedback

The customer has organized on-site acceptance, and the video recognition is accurate and effective. The radar recognition is fast and accurate, and all functions meet the technical requirements. The acceptance is qualified. At the same time, the customer appreciates the team's professional skills demonstrated by the micro technical engineers who overcame the cold environment on site and persisted in working, as well as the backend software personnel who responded quickly and cooperated in a timely manner.

Postscript:

As a lifting safety protection expert, Weite has always focused on the research of lifting operation management, accumulated rich on-site experience in various fields, and is committed to providing users with safer and more efficient customized solutions.