Remote Sensing Target Detection Method Based on Boundary Constrained Centernet
A boundary constraint and target detection technology, which is applied to target detection in optical remote sensing images, image target detection, and target detection based on boundary constraints CenterNet, can solve the problems of low detection accuracy and low recall rate of dense small targets, and achieve the goal of improving Detection accuracy, the effect of improving detection precision and recall rate
Active Publication Date: 2021-09-03
XIDIAN UNIV
View PDF4 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] The purpose of the present invention is to address the deficiencies of the above-mentioned prior art, and propose a target detection method based on boundary constraints CenterNet, which is used to solve the technical problems of low detection accuracy and low recall rate of dense small targets in the prior art
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
[0029] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.
[0030] refer to figure 1 , the implementation steps of the present invention are as follows:
[0031] Step 1) Get the training sample set:
[0032] Randomly select N pieces of images with a pixel size of W×H×D from the optical remote sensing image data set as a training sample set, where N=10000, W=H=511, D=3;
[0033] Step 2) Construct the boundary constraint CenterNet network:
[0034] (2a) Build a feature extraction network, a boundary-constrained convolutional network, and a keypoint generation network, where:
[0035] The feature extraction network includes the first input layer, the first downsampling convolutional layer, the first convolutional layer, the second downsampling convolutional layer, the second convolutional layer, the third downsampling convolutional layer, the Four downsampling convolutional layers, fifth down...
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 proposes a remote sensing target detection method based on boundary constraint CenterNet, which is used to solve the technical problem of low detection accuracy and low recall rate of dense small targets in the prior art. The implementation steps are: obtaining a training sample set; constructing a boundary Constrain the CenterNet network; obtain the predicted label and embedding vector of the training sample set; calculate the loss of the boundary constraint CenterNet network; train the boundary constraint CenterNet network; obtain the target detection result based on the trained boundary constraint CenterNet network. The present invention performs maximum pooling in the constrained pooling area through the corner constrained pooling layer, extracts fine features around the target, effectively improves the detection accuracy and recall rate of dense small targets, and utilizes the boundary generated by the boundary constrained convolution network Constraint labels constrain the prediction frame to obtain a more accurate target prediction frame and further improve the detection accuracy of the target.
Description
technical field [0001] The invention belongs to the technical field of machine vision and relates to an image target detection method, in particular to a boundary-constrained CenterNet-based target detection method, which can be used for target detection in optical remote sensing images. Background technique [0002] The target detection method is one of the core research contents in the field of machine vision. It is a technique for regressing and classifying all the targets of interest in the image by extracting and processing image features, and determining their positions and categories. It is widely used in Object Detection in Optical Remote Sensing Images. The technical indicators of target detection methods include detection accuracy, recall rate, and detection speed. In remote sensing images, affected by the image resolution, there are a large number of dense small targets. The small ratio in the whole image makes it difficult to accurately detect the existence of d...
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): G06K9/00G06K9/62
CPCG06V20/13G06V20/53G06V2201/07G06F18/214
Inventor 冯婕曾德宁李迪焦李成张向荣曹向海刘若辰尚荣华
Owner XIDIAN UNIV