Safety early warning visual detection method for coke tank car
A security early warning and visual detection technology, applied in the field of image target visual detection, can solve problems such as difficult image acquisition, transmission processing, detection and control, harsh working environment of equipment, threat to hardware equipment operation stability, and recognition result reliability. Improved accuracy and robustness, improved accuracy, and high portability
Pending Publication Date: 2019-10-18
天津中科智能识别有限公司
View PDF7 Cites 2 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Due to the complex and changeable actual site conditions, the harsh working environment of equipment, and the widespread existence of high-temperature dust, this will cause great difficulties in image acquisition, transmission, processing, detection and control, and at the same time threaten the stability of hardware equipment operation and the reliability of recognition results.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0023] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in combination with specific examples and with reference to the detailed drawings.
[0024] The coke tank truck safety early warning visual inspection method proposed by the present invention has a flow chart as follows figure 1 shown, including the following steps:
[0025] S1: Annotate the images collected on-site by focusing on the obstacles on the road of the tanker to form an annotated image sub-module;
[0026] S2: Design a single-shot multi-box detector according to the site conditions and collected images, use the marked images in the labeled image sub-module for training, and obtain a deep learning model sub-module based on the single-shot multi-box detector;
[0027] S3: The image processing sub-module uses the camera to obtain the real-time video image of the coke tanker moving, and after preprocessi...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a safety early warning visual detection method for a coke tank car. The method comprises the following steps: marking an acquired image of an obstacle on an advancing path of the coke tank car on site; designing a single-shot multi-box detector, and obtaining a deep learning model sub-module based on the single-shot multi-box detector by using the marked image to perform training; acquiring a real-time video image of the coke tank car in the advancing process, preprocessing the real-time video image, and then sequentially inputting the preprocessed real-time video imageinto the trained deep learning model sub-module based on the single-shot multi-box detector for detection to obtain a feedback obstacle detection result; analyzing a detection result, eliminating a false identification result, and then storing the non-security image; carrying out danger type analysis on the data detected to be qualified by the false alarm processing sub-module; and transmitting adanger type analysis result and a stability monitoring result of factors influencing the operation stability periodically in the synchronization thread to the coke tank car PLC by using a communication protocol so as to control the motion state of the coke tank car. According to the method, the safety monitoring of the working process of the focusing tank car is realized.
Description
technical field [0001] The invention relates to the technical field of image target visual detection, in particular to a coke tank car safety early warning visual detection method for solving the safety warning in the coke tank car driving scene in the coking industry. Background technique [0002] With the development of intelligent technology, the era of industrial intelligence has arrived, and the intelligentization of traditional industrial production is also imperative. The working environment of the iron and steel industry is harsh, and the operation control of many automation equipment needs intelligent transformation. Visual inspection based on deep learning is to use machines instead of human eyes to do automatic detection and judgment to control the execution of on-site equipment. It can replace or assist humans in the process of real-time perception and intelligent decision-making to a certain extent, and at the same time, it has automation. The advantages of rel...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/62
CPCG06V20/48G06V20/58G06F18/214
Inventor 张堃博申振腾杨程午孙哲南
Owner 天津中科智能识别有限公司