A method for identifying open flames in cable tunnels based on ssd
A recognition method and cable tunnel technology, applied in neural learning methods, character and pattern recognition, instruments, etc., can solve the problems of inability to control the fire scene and real-time scheduling, and achieve strong anti-interference ability, high accuracy and stability. Good results
Active Publication Date: 2020-02-21
GUANGDONG POWER GRID CO LTD +1
View PDF5 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, traditional sensors have defects such as being unable to control the fire scene, and unable to perform real-time scheduling according to the actual situation.
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
no. 1 example
[0029] This embodiment is the first embodiment of the SSD-based cable tunnel open flame identification method of the present invention, including the following steps:
[0030] S1. Capture and collect tunnel open flame sample images through the tunnel inspection robot camera to obtain an image set;
[0031] S2. Traversing all sample images in the image set described in step S1, enclosing a complete open flame area with a bounding box for each sample image and marking it as an open flame sample image with a bounding box, marking the pixels in the bounding box as Open flame category, mark the pixels outside the bounding box as the background category, and obtain the first training set;
[0032] S3. Carry out scaling processing for the open flame sample images with bounding boxes described in step S2, transform the longer side of the bounding box of each open fire sample image to the preset target size, and transform the shorter side to the longer side according to the longer side...
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 present invention relates to the technical field of computer image processing and recognition methods, and more specifically, relates to an SSD-based cable tunnel open flame recognition method. The method of training the SSD network model is selected, and the SSD network is fully utilized to select feature maps of different levels for multiple The characteristics of scale detection make up for the shortcomings of existing methods that are less robust to small target object detection. The invention can accurately detect and locate the open flame area in the image and complete the accurate result detection, so that the placement position of the camera and the fixed-point inspection position of the inspection robot are more free, and can effectively deal with some complex environmental conditions. The invention has the advantages of convolutional neural network, which can reduce the interference of factors such as complex image background and dim tunnel environment during the open flame detection process taken by the robot inside the tunnel. It has strong anti-interference ability and improves the accuracy of tunnel open flame detection.
Description
technical field [0001] The present invention relates to the technical field of computer image processing and identification methods, and more particularly, relates to an SSD-based identification method for open flames in cable tunnels. Background technique [0002] In recent years, cable tunnels have been developed on a large scale. However, due to the long and narrow space of the cable tunnel, fewer entrances and exits, and dense high-voltage cables, once a fire accident such as cable burning occurs, it is difficult to implement fire fighting measures. At the same time, once the fire spreads, it will cause great harm to the electrical equipment in the tunnel and the safety of the city. At present, traditional fire detection in cable tunnels generally uses sensors such as temperature sensing, smoke sensing, and photoelectric sensors. However, traditional sensors have shortcomings such as being unable to control the fire scene, and unable to perform real-time scheduling acc...
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 Patents(China)
IPC IPC(8): G06K9/00G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V20/40G06N3/045G06F18/24147
Inventor 裴星宇黄顺涛崔江静仇炜黄培专朱五洲袁永毅廖雁群刘淡冰
Owner GUANGDONG POWER GRID CO LTD