Automatic inspection image style transfer display method
A technology of automatic inspection and display method, which is applied in image data processing, graphic image conversion, digital data information retrieval, etc.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0040] An automatic inspection image style migration display method (see attached figure 2 ), which is applicable to the preprocessing of image acquisition during the automatic inspection of electric power equipment, and the automatic inspection of electric power equipment adopts an automatic robot for inspection, including the following steps:
[0041] Step 1, obtaining the real-time image of the target device to form a content picture for collection and communication transmission;
[0042] The content image in this embodiment is manually set, mainly based on the appearance of the device, supplemented by the main detection data of the device that needs to be displayed directly.
[0043] The shooting position of the inspection is set with a two-dimensional code or other content recording method that enables the automatic robot to realize shooting according to predetermined parameters. In this embodiment, the inspection method of the automatic robot includes magnetic navigatio...
Embodiment 2
[0080] This embodiment is basically the same as Embodiment 1. The difference is that in this embodiment, the neural network is scaled to a certain ratio of depth and width, and the width of the convolutional neural network is the same as the number of filters in the convolutional layer. Associated, the resolution is the size of the input image, and the depth is increased by stacking more convolutional layers. In general, only the depth and width of the network are adjusted according to the set ratio. Only when the adjustment amount of the network width and depth exceeds the threshold, the neural network model also adjusts the resolution coefficient of the input image accordingly. Scaling the width of the network allows layers to learn more fine-grained features. However, the improvement of accuracy is not directly proportional to the width and depth of the zoom network. Therefore, it is estimated that the resource utilization rate of each layer is selected to select the approp...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com