Pomacea canaliculata egg detection method based on multi-scale feature fusion and dynamic convolution

A multi-scale feature and detection method technology, applied in the field of computer vision, can solve the problems of low recognition accuracy of apple snail eggs and poor model robustness, and achieve the effect of improving accuracy, effectively extracting features, and enhancing recognition ability
CN113642410APending Publication Date: 2021-11-12NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Current Assignee / Owner
NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
Publication Date
2021-11-12

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses a pomacea canaliculata egg detection method based on multi-scale feature fusion and dynamic convolution, belongs to the technical field of computer vision, and improves the accuracy of current pomacea canaliculata egg detection. The method comprises the following steps: firstly, collecting aerial images of ampullaria gigas eggs, and marking the eggs in the aerial images; then on the basis of the darknet53 network structure, using four-scale feature fusion and dynamic convolution to construct a new network structure; sending the obtained pomacea canaliculata egg data set into a neural network for training until the network converges, and obtaining a weight file; and then detecting the ampullaria gigas egg target in the test image by using the trained neural network and the weight file, and outputting a detection result. According to the invention, the problems that the current ampullaria gigas egg recognition accuracy is low, and the model has a low ampullaria gigas egg recognition rate in a real natural environment are solved.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention belongs to the technical field of computer vision, and in particular relates to a method for detecting apple snail eggs based on multi-scale feature fusion and dynamic convolution. Background technique

[0002] The rapid development of computer vision in recent years has made scenarios such as smart agriculture and scientific epidemic prevention a reality. People are increasingly inclined to use computer vision technology to detect potential pest hazards. Most existing object detection methods can only classify eggs, or detect them in specific background conditions rather than natural environments, and they cannot achieve good results in complex real-world scenes.

[0003] In terms of pest control, in order to solve the classification problem, Konantinos P. Ferentinos et al. developed a special deep learning model based on the VGG convolutional neural network architecture to identify plant diseases from simple images of healthy or disease...

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