A Scene Character Recognition Method Based on Continuous Convolution Activation
A character recognition and convolution technology, which is applied in the field of scene character recognition based on continuous convolution activation, can solve the problems of not being able to fully retain feature information and stroke structure information, affecting the accuracy of scene character recognition, etc., so as to improve the accuracy. Effect
Active Publication Date: 2021-10-15
晶程甲宇科技(上海)有限公司
View PDF7 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Although the above methods have achieved greater success, they only use the convolutional activation map in a single convolutional layer for feature representation and ignore the information provided by other convolutional layers, so they cannot fully retain significant feature information and stroke structure. information, thus affecting the accuracy of scene character recognition
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0047] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in combination with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are exemplary only, and are not intended to limit the scope of the present invention. Also, in the following description, descriptions of well-known structures and techniques are omitted to avoid unnecessarily obscuring the concept of the present invention.
[0048] figure 1 It is a flowchart of a scene character recognition method based on continuous convolution activation according to an embodiment of the present invention, as follows figure 1 Some specific implementation processes of the present invention are described by way of example. The present invention is based on the scene character recognition method of continuous convolution activation and comprises the followin...
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
Login to View More Abstract
The embodiment of the present invention discloses a scene character recognition method based on continuous convolution activation. The method includes: inputting the training image into the convolution neural network to obtain the convolution activation map; using the convolution activation map in the first preset convolution layer The convolutional activation map is obtained to obtain the convolutional activation descriptor; the convolutional activation map in the second preset convolutional layer is used to obtain the weight matrix; based on the convolutional activation descriptor and the weight matrix, the continuous convolutional activation descriptor is obtained; using Fisher The vector encodes the continuous convolution activation descriptor to obtain the feature vector of the training image; based on the feature vector, the scene character recognition classification model is obtained by using the support vector machine; the feature vector of the test image is obtained, and input to the scene character recognition classification model to obtain the scene character recognition result. The present invention combines bottom-level feature information such as strokes and textures with high-level semantic information in feature vectors to achieve the purpose of effectively mining salient feature information and stroke structure information and improving the accuracy of scene character recognition.
Description
technical field [0001] The invention belongs to the technical fields of pattern recognition and artificial intelligence, and in particular relates to a scene character recognition method based on continuous convolution activation. Background technique [0002] Characters as a medium of image communication are ubiquitous in practical applications and provide valuable semantic cues for various applications such as automatic geocoding, product search, robot navigation, and image and video retrieval. Scene characters are characters that appear in real scene images, and they are easily disturbed by various factors, such as non-uniform illumination, complex background, font distortion, blur, font changes, etc. Therefore, accurately recognizing scene characters is a particularly challenging task. [0003] In the past few decades, scene character recognition has become a research hotspot, and researchers have proposed many scene character recognition algorithms. Some early methods...
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
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/32G06K9/62
CPCG06V20/63G06V30/10G06F18/2132G06F18/2411G06F18/2413
Inventor 张重王红刘爽
Owner 晶程甲宇科技(上海)有限公司