Unlock instant, AI-driven research and patent intelligence for your innovation.

SAR image sea surface ship detection method based on yoov3 algorithm and sliding window strategy

A sliding window and ship detection technology, which is applied in the field of computer vision, can solve the problems of large ship traffic and the inability to visually display the supervision of ships on the sea surface, etc., and achieve the effect of fast recognition speed and high detection accuracy

Active Publication Date: 2020-06-05
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF6 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, due to the vast sea area of ​​our country and the huge flow of ships, the supervision of ships on the sea cannot be displayed intuitively.

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • SAR image sea surface ship detection method based on yoov3 algorithm and sliding window strategy
  • SAR image sea surface ship detection method based on yoov3 algorithm and sliding window strategy
  • SAR image sea surface ship detection method based on yoov3 algorithm and sliding window strategy

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0036] The SAR image sea ship detection method based on the yolov3 algorithm and the sliding window strategy is characterized in that it includes the following steps: S1, data preprocessing: the SAR image data is marked and segmented, and the VOC2012 image data set is subjected to three-channel graying Degree balance processing; S2, on the basis of step S1, carry out yolov3 model pre-training: use VOC2012 picture data set as input data, put into general yolov3 model and carry out model training; S3, on the basis of step S2, after training yolov3 model for migration learning: put SAR image data into the pre-trained yolov3 model for optimization, and obtain the target yolov3 detection model; S4, on the basis of step S3, perform sliding window strategy processing on real-time SAR image data, and then input to The target yolov3 detection model is used to obtain detection results, and a clustering algorithm is performed on the detection results to deduplicate them.

[0037] Prefera...

Embodiment 2

[0062] On the basis of Embodiment 1, the present invention mainly includes two parts, one is the training of the detection model used in the bottom layer, and the other is the combination of the detection model + sliding window detection strategy + clustering algorithm to deduplicate Methods and methods for detecting ships on the sea surface. The first part mainly includes the following steps 1 and 2; the second part mainly includes the following steps 3 and 4.

[0063] Step 1: Preprocess the data. The data set contains two parts: the SAR image data of Gaofen-3 with a resolution of 10 meters and the general data set of VOC2012. Firstly, segment the SAR image data of Gaofen-3 with 10-meter resolution to obtain several original data images with a size of 256X256 pixels, and then use the LabelImg tool written in python language to manually label the obtained data set to obtain the Corresponding position information in the picture, the xml format file of the picture information w...

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

No PUM Login to View More

Abstract

The invention discloses an SAR image sea surface ship detection method based on a yolov3 algorithm and a sliding window strategy, and the method comprises the steps: data preprocessing: carrying out the data labeling and segmentation of SAR image data, and carrying out the three-channel gray balance processing of a VOC2012 image data set; pre-training a yolov3 model: taking the VOC2012 picture data set as input data, and putting the VOC2012 picture data set into a universal yolov3 model to carry out model training; performing transfer learning on the trained yolov3 model; inputting SAR image data into the pre-trained yolov3 model for optimization to obtain a target yolov3 detection model; and performing sliding window strategy processing on the real-time SAR image data, and then inputtingthe processed real-time SAR image data into the target yolov3 detection model to obtain a detection result, and performing clustering algorithm deduplication on the detection result. According to themethod, a good detection effect can be achieved on the low-resolution SAR image, the detection speed can be adjusted, the method has the advantages of being high in recognition speed and high in detection precision, and the method can also be used for target detection of other types of satellite images.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a SAR image sea ship detection method based on a yolov3 algorithm and a sliding window strategy. Background technique [0002] A ship is a means of transportation that can sail or berth in waters for transportation or operations. my country is a large maritime country. There are a large number of ships operating in my country's territorial waters every day. The management and monitoring of ships is an important responsibility of the maritime department. However, due to the vast sea area of ​​our country and the huge flow of ships, the supervision of ships on the sea cannot be displayed intuitively. SAR (Synthetic Aperture Radar), that is, synthetic aperture radar, is an active earth observation system that can be installed on aircraft, satellites, spacecraft and other flight platforms to observe the earth all day and all weather, and has a certain Surface pen...

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
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/32G06N3/04
CPCG06V20/13G06V10/25G06N3/045
Inventor 郑泽忠江邵斌刘佳玺李锴牟范侯安锴李江
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA