Traffic sign recognizing method based on multi-resolution convolution neural networks
A convolutional neural network and traffic sign recognition technology, applied in character and pattern recognition, instruments, computer components, etc., can solve the problem of not being able to balance speed and accuracy, and achieve the effect of improving training speed
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0046]
[0047] 1, determine training set, what the present invention selects is the training set in GTSRB (Germany traffic sign recognition benchmark, German traffic sign recognition benchmark), comprises training picture 39,209 pieces, test picture 12630 pieces.
[0048] 2. Preprocess the pictures in the training set. The steps are to first process them into grayscale images, and then normalize them into pictures with a uniform resolution of 48*48, and then back up these pictures to cut out the central area. A cropped image with a resolution of 36*36, attached figure 1 Input samples for two different resolutions of the example, and start training with these two parts of the image as two inputs.
[0049] 3. Select a training group; randomly select 50 samples each time from the sample set as a training group.
[0050] 4. Put each weight v ij ,w jk and threshold θ k , set to a small random value close to 0, and initialize the precision control parameters ε, learning r...
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