An image recognition method of sprouted potato based on improved yolov5 model
A technology of image recognition and potato, which is applied in the field of image recognition, can solve the problems of speed reduction and achieve the effects of improving accuracy, improving the ability to identify germinated potatoes, and improving the generalization ability
Active Publication Date: 2022-07-19
AGRI INFORMATION INST OF CAS
View PDF8 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
The mAP of the four models increases sequentially, but the speed decreases sequentially
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
[0052] The invention provides a sprouting potato image recognition method based on the improved yolov5 model, comprising the following steps:
[0053] S1, construct an image recognition network model based on the improved yolov5 model, the image recognition network model specifically includes an image tensor of 640*640 input layer, Backbone Network, Neck Network and Detector, wherein, the input layer adds mosaic (Mosaic) Data enhancement module; the Backbone Network is the core structure of YOLOv5, consisting of Focus, Conv, C3 (CSP Bottleneck with 3 convolutions) and Spatial Pyramid Pooling (SPP) modules; Interval slices are then stitched together. Compared with convolution downsampling, the output depth of Focus is increased by 4 times and more image information is retained. Conv is the basic convolution unit of YOLOv5, which sequentially performs two-dimensional convolution, regularization, and activation operations on the input. C3 is composed of several Bottleneck module...
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 method for sprouting potato image recognition based on an improved yolov5 model, comprising the following steps: constructing an image recognition network model, acquiring a potato image data set, and dividing it into a training set and a test data set; performing preprocessing analysis on the data set, Obtain the data visualization results, and determine the model used in the feature fusion of the data set based on the visualization results; use the training set to train the image recognition network model, obtain various parameter indicators of the image recognition network model, and obtain the trained image recognition network model; The trained image recognition network model is tested using the test data set, and the accuracy of the test results is evaluated. Use the 9‑Mosaic algorithm to enhance and optimize the obtained potato image before entering the backbone, strengthen the small target features, adjust the transfer rules of the intra-graph convolution in the backbone, and use the cross-graph convolution CrossConv instead of Conv to connect two intra-graph convolutions to enhance Feature similarity improves the accuracy of sprouting potato identification.
Description
technical field [0001] The invention relates to the technical field of image recognition, in particular to an image recognition method of sprouted potato based on an improved yolov5 model. Background technique [0002] Potatoes are at the heart of the world's food supply and are grown in more than 100 countries. As the fourth largest crop after maize, wheat and rice, potato production is a major concern for the food industry, supporting numerous research projects. In particular, an important process in the potato industry, the storage of potatoes, is regularly studied [1] . China's strategy of using potatoes as a staple food clearly shows the need to improve potato processing, and the bottleneck of this strategy lies in the selection of suitable raw materials and the technology and equipment for processing potatoes. In fact, the germination of potatoes during the storage period during processing is very deadly to the entire industry. The sprouted potatoes contain only 0.2...
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): G06V10/764G06V10/774G06V10/80G06K9/62G06N3/04G06N3/08
CPCG06N3/086G06N3/045G06F18/253G06F18/24G06F18/214
Inventor 代国威闫燊胡林樊景超张翔鹤
Owner AGRI INFORMATION INST OF CAS