Training method for improving image recognition accuracy with assistance of electronic tag recognition
A technology of electronic tags and training methods, applied in the field of image recognition, can solve the problem of high cost, achieve the effect of improving accuracy and avoiding excessive cost
Active Publication Date: 2020-05-15
EAST CHINA UNIV OF SCI & TECH +1
View PDF6 Cites 2 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0010] In order to overcome the defects of the high cost of existing RFID and the high cost of retraining the model when the commodity category changes in the image recognition technology based on neural network, the present invention provides an electronic tag recognition system that assists in improving the accuracy of image recognition. training method
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
[0031] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.
[0032] In the present invention, the SSD neural network algorithm refers to the SSD (single shotmultibox detector) algorithm applying CNN (neural network), which is a conventional technology in the art, and it also involves adjusting the loss function and convolution parameters according to the training results , are conventional technical means in the field, so in the following embodiments, the algorithm itself will not be described in detail.
[0033] like figure 1 As shown, the training method for electronic tag recognition in this embodiment to assist in improving the accuracy of image recognition includes steps:
[0034] S1. Collect images and perform preprocessing to construct training sets and test sets; wherein, the background of the image should have a larger color difference than the commodity to be identified. Preferably, the prepro...
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 a training method for improving image recognition accuracy with the assistance of electronic tag recognition, and the method comprises the steps: carrying out the preliminary training and testing of a detection model for image recognition, enabling the image recognition accuracy of the detection model to reach a first qualified threshold, and obtaining an intermediate detection model; identifying the changed to-be-detected commodity through the electronic tag and the intermediate detection model at the same time: when the two identification results are inconsistent, updating the intermediate detection model by using the information identified by the electronic tag and the image information before and after the change of the commodity; repeating the process and counting the identification accuracy of the intermediate detection model until the identification accuracy reaches a second qualified threshold, thereby obtaining an optimized detection model. The optimized detection model can be packaged into a callable program module, an electronic tag does not need to be arranged in the using process any more, and the detection model has good reliability in a commodity change scene.
Description
technical field [0001] The invention belongs to the field of image recognition, and specifically relates to a training method for electronic tag recognition to assist in improving the accuracy of image recognition. Background technique [0002] At present, the unmanned retail industry is extremely hot, and the key technology involved is the identification of retail goods. Today, with high labor costs, automatic commodity identification technology is one of the important technologies to ensure the efficient circulation of commodities, and there is a great demand in unmanned retail scenarios. [0003] Existing commodity automatic identification technologies include barcode identification technology, RFID (radio frequency identification tag), computer image recognition technology, etc. [0004] Among them, barcode recognition technology is the most mature product recognition technology at this stage, but there are still many shortcomings: [0005] 1) It is usually necessary t...
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): G06K9/18G06K9/62G06N3/04G06N3/08G06V30/224
CPCG06N3/08G06V30/224G06N3/045G06F18/241
Inventor 徐坚强胡注娇
Owner EAST CHINA UNIV OF SCI & TECH