![一种基于卷积神经网络的鸟类识别方法和装置](/CN/2019/1/175/images/201910875685.jpg)
基本信息:
- 专利标题: 一种基于卷积神经网络的鸟类识别方法和装置
- 专利标题(英):Bird identification method and device based on convolutional neural network
- 申请号:CN201910875685.7 申请日:2019-09-17
- 公开(公告)号:CN110837771A 公开(公告)日:2020-02-25
- 发明人: 华斯亮 , 徐健 , 卞九辉 , 华强 , 张静亚 , 刘玉申 , 洪学鹍
- 申请人: 常熟理工学院
- 申请人地址: 江苏省苏州市常熟市南三环路99号
- 专利权人: 常熟理工学院
- 当前专利权人: 常熟理工学院
- 当前专利权人地址: 江苏省苏州市常熟市南三环路99号
- 代理机构: 南京苏高专利商标事务所
- 代理人: 吴静波
- 主分类号: G06K9/00
- IPC分类号: G06K9/00 ; G06K9/34 ; G06K9/62 ; G06N3/04 ; G06N3/08
The invention discloses a bird identification method based on a convolutional neural network, which is applied to one or more pieces of electronic equipment, and comprises the following steps: (1) acquiring an original picture by an acquisition end; (2) enabling the identification end to receive the photo; (3) preprocessing the photo received in the step (2), specifically, the optimized convolutional neural network MobileNetv2 performs segmentation, feature extraction, feature selection and feature classification with overlapped regions on the image according to pixels of the original image; (4) enabling the optimized target recognition model to recognize the preprocessed images, splicing the recognition results into the positions of the original images according to the cutting size and step length, and performing fusion processing on the recognition results of the overlapping areas of the cut images; (5) comparing the data sets of the bird images to identify, store and count birds; for the pictures which do not identify birds, the pictures can be discarded or reserved or partially reserved according to the storage configuration of the identification end. According to the method provided by the invention, the identification precision of the picture can be effectively improved.
IPC结构图谱:
G | 物理 |
--G06 | 计算;推算;计数 |
----G06K | 数据识别;数据表示;记录载体;记录载体的处理 |
------G06K9/00 | 用于阅读或识别印刷或书写字符或者用于识别图形,例如,指纹的方法或装置 |