Electronic map adaptive classification method, device, equipment and storage medium
An adaptive classification and electronic map technology, applied in the field of electronic maps and computer vision, can solve the problems of large number of maps, low classification accuracy, and low efficiency, and achieve efficient and accurate automatic identification and classification
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0026] figure 1 It is a flow chart of the electronic map adaptive classification method provided by Embodiment 1 of the present invention. This embodiment is applicable to the situation of reviewing electronic maps on the Internet. This method can be executed by an electronic map adaptive classification device. The device It can be realized by means of software and / or hardware, and integrated in computer equipment.
[0027] Such as figure 1 As shown, the electronic map adaptive classification method in the embodiment of the present invention specifically includes:
[0028] S101. Obtain electronic map data to be classified.
[0029] In recent years, "problem map" incidents have occurred from time to time. It often occurs that my country's important islands and national boundaries are wrongly drawn on the map, or sensitive and confidential information is marked on the map. These will endanger national sovereignty, national security and hurt national sentiment. Therefore, it i...
Embodiment 2
[0036] Figure 2a A schematic flow chart of the neural network classifier training method provided in Embodiment 2 of the present invention, which is used to train the neural network classifier, such as Figure 2a Shown, described neural network classifier training method comprises:
[0037] S201. Acquire a target convolutional neural network model, where the target convolutional neural network model includes a plurality of sequentially arranged convolutional neural networks.
[0038] The target convolutional neural network model is constructed from multiple different convolutional neural networks based on the model fusion method. Exemplarily, the target convolutional neural network model includes 3 different convolutional neural networks, and the arrangement is ResNet50, Xception, InceptionV3.
[0039] S202. Based on the target convolutional neural network model, extract the features of the sample data in the training set and save them as feature vectors, wherein the number...
Embodiment 3
[0050] image 3 It is a schematic flowchart of the neural network classifier training method provided by the third embodiment of the present invention. In this embodiment, on the basis of the above-mentioned embodiments, the acquisition of the target convolutional neural network model is further optimized, such as image 3 Shown, described neural network classifier training method comprises:
[0051] S301. Build different convolutional neural networks based on multiple deep learning frameworks.
[0052] Build a convolutional neural network with good performance on the deep learning framework. Specifically, use the championship framework InceptionV3, ResNet50, Xception, InceptionResNetV2, VGG19 and VGG16 in ILSVRC (ImageNet Large Scale Visual Recognition Challenge) to build 6 different convolutions Neural Networks.
[0053] S302. Cut off the last fully connected layer of the different convolutional neural networks, and replace the fully connected layer with a global average ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More