The invention discloses an image recognition module based on a deep neural network
A technology of deep neural network and image recognition, which is applied in the direction of biological neural network model, neural architecture, character and pattern recognition, etc. It can solve the problems that there is no integrated image recognition module for integrated image collection, so as to improve the generalization ability, Low-latency, high-efficiency effects
Pending Publication Date: 2019-05-31
EAST CHINA NORMAL UNIV
View PDF11 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
At present, there is no report on an integrated image recognition module that integrates image acquisition, training and 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 1
[0015] See attached figure 2 , the present invention performs the image recognition operation in this way: the camera module 2 inputs the acquired target image into the arithmetic controller 1 and displays it on the screen by the display module 3, and the acquired target image is uploaded after being marked and classified by the arithmetic controller 1. Go to the cloud server, perform model training based on the acceleration algorithm of the pre-trained model through the cloud image training module 4, and download the model parameter file after the training to the operation control module 1 through the local server to achieve end-to-end target image acquisition, training and recognition. Integrated, the results are returned to the display module 3 for screen display, and the results are output and fed back to the user through wired or wireless means. The cloud image training module 4 is based on the deep neural network model, and accelerates the new image training through the...
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 invention discloses an image recognition module based on a deep neural network. The method is characterized in that an operation control module is adopted; Camera module, the image recognition module consists of a display module and a cloud image training module; a target image acquired by the camera module is marked and classified by the operation control module and then is uploaded to the cloud image training module; and the cloud image training module utilizes a deep neural network to train the uploaded image and marked and classified data, a trained model is downloaded to the operationcontrol module, and the end-to-end target image is acquired, trained and identified. Compared with the prior art, the device has an end-to-end function; The method has the advantages of simple structure, low cost, high efficiency and low time delay, the whole process from image acquisition to image training and image recognition is completely completed based on an end-to-end module, a user does not need to pay attention to an image recognition algorithm and implementation thereof, the structure is simple, operation is convenient, adjustment can be carried out according to requirements, and the application requirements of different scenes are met.
Description
technical field [0001] The invention relates to the technical field of image recognition, in particular to an end-to-end image recognition module based on a deep neural network. Background technique [0002] The deep neural network has unique advantages in the field of image recognition. It can imitate the structure of the human eye and can recognize the visual field efficiently. Compared with the traditional BP neural network, it has the advantages of fewer parameters and higher recognition efficiency. Image recognition is a branch of machine vision, which requires accurate image recognition results in applications and also has certain requirements for delay. Subsequent actuators, such as robots, trolleys, etc., need to perform decision-making control based on the results of image recognition. [0003] At present, deep neural network technology has been widely used in engineering, pattern recognition, industrial robots, machine vision, etc. In view of the huge market of i...
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 Applications(China)
IPC IPC(8): G06K9/62G06N3/04
Inventor 陈康胡孟晗李庆利
Owner EAST CHINA NORMAL UNIV