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

Fast Object Detection Method Based on Lightweight Neural Network

A technology of target detection and neural network, which is applied in the field of fast target detection based on lightweight neural network, can solve the problems of large size of neural network, consumption of storage space, time-consuming and inefficient manual extraction of image features, and slow detection speed, etc. The effect of fast speed, fast detection speed and improved accuracy

Active Publication Date: 2021-07-13
NANJING UNIV OF SCI & TECH
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The present invention aims at the problems of time-consuming and inefficient manual extraction of image features in the above-mentioned target detection method, the large volume of the neural network in the deep learning method consumes storage space, the complexity of the network model, and the slow detection speed, and proposes a lightweight neural network-based Fast Object Detection Method

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
  • Fast Object Detection Method Based on Lightweight Neural Network
  • Fast Object Detection Method Based on Lightweight Neural Network
  • Fast Object Detection Method Based on Lightweight Neural Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0047] The present invention will be further described below in conjunction with accompanying drawing.

[0048] The target detection method of the present invention includes three main processes of neural network construction, neural network training and target detection.

[0049] The neural network construction process refers to designing a reasonable number of neural network layers, using the optimal number of convolution kernels in each layer of the neural network, and encapsulating the modules with multiple layers of convolutional layers and pooling layers that are reused. Include the following specific steps:

[0050] First, design the front part of the network. The front part is the first few layers of the convolutional neural network, which is responsible for extracting the basic features of the input image and encapsulating it as a Front module, such as figure 1 shown. The Front module consists of three convolutional layers and one pooling layer. Three convolutional...

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 a fast object detection method based on a lightweight neural network. Including building convolutional neural network, training convolutional neural network and target detection: defining the main modules of convolutional neural network Front module and Tinier module; defining the number of layers of convolutional neural network and the position of pooling layer; preprocessing the data set, Convert it into the standard format of the target detection framework Darknet input; initialize the parameters of the convolutional neural network; train the neural network model through continuous iterative forward propagation and back propagation; input the test image, and use the neural network model obtained from the training process to calculate Obtain the detection numerical results; mark the image according to the detection results, and mark the position and category of each object with a rectangular frame. The detection speed using the method of the invention is faster and the accuracy rate is also higher.

Description

technical field [0001] The invention relates to the fields of pattern recognition and video analysis, more specifically, a fast target detection method based on a lightweight neural network. Background technique [0002] Object detection is an important research topic in the field of computer vision. It has been widely used in multiple real-world applications, such as face recognition, traffic safety, crowd monitoring and image retrieval. Real-time target detection based on deep learning refers to marking the position and category of target objects in a natural scene picture or video. In the face of massive image and video data, manual labeling is time-consuming and inefficient, and automatic and fast object detection methods are urgently needed. [0003] The object detection method based on deep learning contains two key steps: feature extraction and classifier (regressor) training. Unlike the traditional method of manually extracting the features of the target object, 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
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/62G06N3/04G06N3/08
CPCG06N3/084G06N3/045G06F18/214
Inventor 刘亚洲曹森
Owner NANJING UNIV OF SCI & TECH