CN112184322A - 一种基于图片识别的活体禽畜抵押贷款贷前估值方法 - Google Patents

一种基于图片识别的活体禽畜抵押贷款贷前估值方法 Download PDF

Info

Publication number
CN112184322A
CN112184322A CN202011091917.9A CN202011091917A CN112184322A CN 112184322 A CN112184322 A CN 112184322A CN 202011091917 A CN202011091917 A CN 202011091917A CN 112184322 A CN112184322 A CN 112184322A
Authority
CN
China
Prior art keywords
livestock
poultry
training
yolov4
loan
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202011091917.9A
Other languages
English (en)
Inventor
林闽
唐雨晴
杨帆
龚泠方
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Agricultural Bank of China Sichuan Branch
Original Assignee
Agricultural Bank of China Sichuan Branch
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Agricultural Bank of China Sichuan Branch filed Critical Agricultural Bank of China Sichuan Branch
Priority to CN202011091917.9A priority Critical patent/CN112184322A/zh
Publication of CN112184322A publication Critical patent/CN112184322A/zh
Pending legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00Commerce
    • G06Q30/02Marketing; Price estimation or determination; Fundraising
    • G06Q30/0278Product appraisal
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/21Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/04Architecture, e.g. interconnection topology
    • G06N3/045Combinations of networks
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/03Credit; Loans; Processing thereof

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • General Physics & Mathematics (AREA)
  • Development Economics (AREA)
  • Strategic Management (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Economics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Technology Law (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Game Theory and Decision Science (AREA)
  • Image Analysis (AREA)

Abstract

本发明公开了一种基于图片识别的活体禽畜抵押贷款贷前估值方法,包括如下步骤:建立优质畜禽图片数据库;获取图片数据库内家禽类、畜牧类图片数据,建立图片数据集;将数据集内图片进行标注处理;将生成带标注的图片数据集划分为训练集和测试集;训练集作为训练模型输入,导出离线训练yolov4动物数量统计模型;测试集作为学习模型输入,调整参数,得到最优的yolov4动物数量统计模型;采集待估值养殖场内畜禽照片;将待估值养殖场内畜禽照片输入最优的yolov4动物数量统计模型中;畜禽价值评估。本发明的有益效果:解决了因缺乏统一质量评估标准规范,对禽畜数量及价值估算不准确的问题。

Description

一种基于图片识别的活体禽畜抵押贷款贷前估值方法
技术领域
本发明涉及活体禽畜抵押贷款估值技术领域,具体的,涉及一种基于图片识别的活体禽畜抵押贷款贷前估值方法。
背景技术
目前现有贷前调查中对抵质押物进行调查的手段,一般有几种,一是直接的现场调研,通过和贷款人现场会谈、实地考察,查看抵质押活体畜禽的真实情况;二是委托调查,通过中介机构或银行自身网络开展调查。针对活体畜禽抵押贷款,金融机构会对借款人的抵押物,也就是活体畜禽进行价值评估,这个一般是贷前调查人员现场到养殖企业或者农户家进行核查。贷前调查人员到现场后,根据自己的经验,对活体畜禽的数量、质量、总价值进行评估。
但是活体禽畜抵押贷款属于新兴业务,对贷前调查人员要求较高,且由于没有统一的质量评估标准规范,对于活体禽畜的数量及总价值的估算可能存在误差。
发明内容
本发明的目的在于,针对上述问题,提出一种基于图片识别的活体禽畜抵押贷款贷前估值方法。
一种基于图片识别的活体禽畜抵押贷款贷前估值方法,包括如下步骤:
S1:建立优质畜禽图片数据库;
S2:获取图片数据库内家禽类、畜牧类图片数据,建立图片数据集;
S3:将数据集内图片进行标注处理;
S4:将生成带标注的图片数据集划分为训练集和测试集;
S5:训练集作为训练模型输入,导出离线训练yolov4动物数量统计模型;
S6:测试集作为学习模型输入,调整参数,得到最优的yolov4动物数量统计模型;
S7:采集待估值养殖场内畜禽照片;
S8:将待估值养殖场内畜禽照片输入最优的yolov4动物数量统计模型中;
S9:畜禽价值评估。
优选的,所述步骤S3中需对数据集中的每一张图片进行标注,在图片中框选目标区域,并标注框选类别。
优选的,所述步骤S4中图片数据集按照8:2的比例随机划分训练集和测试集。
优选的,所述步骤S5中离线训练yolov4动物数量统计模型利用卷积神经网络(CNN),同时提取适用于大多数模型、任务和数据集的通用特征。
优选的,所述通用特征的处理包括:使用加权残差连接(WRC)、阶段部分连接(CSP)、交叉小批量标准化(CmB N)、自对抗训练(SAT)、Mish激活、马赛克数据增强、CmBN、DropBlock正则化和CIoU丢失。
本发明的有益效果:通过建立训练集和测试集,构建最优yolov4动物数量统计模型,对待估值养殖场内禽畜进行优质禽畜数量统计并进行估值;解决了因缺乏统一质量评估标准规范,对禽畜数量及价值估算不准确的问题。
附图说明
图1为本发明的具体实现流程。
具体实施方式
下面结合附图对本发明做进一步的描述。
如图1所示,一种基于图片识别的活体禽畜抵押贷款贷前估值方法,包括如下步骤:
S1:建立优质畜禽图片数据库;
S2:获取图片数据库内家禽类、畜牧类图片数据,建立图片数据集;
S3:将数据集内图片进行标注处理;
S4:将生成带标注的图片数据集划分为训练集和测试集;
S5:训练集作为训练模型输入,导出离线训练yolov4动物数量统计模型;
S6:测试集作为学习模型输入,调整参数,得到最优的yolov4动物数量统计模型;
S7:采集待估值养殖场内畜禽照片;
S8:将待估值养殖场内畜禽照片输入最优的yolov4动物数量统计模型中;
S9:畜禽价值评估。
需要理解的是,所述步骤S3中需对数据集中的每一张图片进行标注,在图片中框选目标区域,并标注框选类别。
需要理解的是,所述步骤S4中图片数据集按照8:2的比例随机划分训练集和测试集。
需要理解的是,所述步骤S5中离线训练yolov4动物数量统计模型利用卷积神经网络(CNN),同时提取适用于大多数模型、任务和数据集的通用特征。
需要理解的是,所述通用特征的处理包括:使用加权残差连接(WRC)、阶段部分连接(CSP)、交叉小批量标准化(CmB N)、自对抗训练(SAT)、Mish激活、马赛克数据增强、CmBN、DropBlock正则化和CIoU丢失。

Claims (5)

1.一种基于图片识别的活体禽畜抵押贷款贷前估值方法,其特征在于,包括如下步骤:
S1:建立优质畜禽图片数据库;
S2:获取图片数据库内家禽类、畜牧类图片数据,建立图片数据集;
S3:将数据集内图片进行标注处理;
S4:将生成带标注的图片数据集划分为训练集和测试集;
S5:训练集作为训练模型输入,导出离线训练yolov4动物数量统计模型;
S6:测试集作为学习模型输入,调整参数,得到最优的yolov4动物数量统计模型;
S7:采集待估值养殖场内畜禽照片;
S8:将待估值养殖场内畜禽照片输入最优的yolov4动物数量统计模型中;
S9:畜禽价值评估。
2.如权利要求1所述一种基于图片识别的活体禽畜抵押贷款贷前估值方法,其特征在于,所述步骤S3中需对数据集中的每一张图片进行标注,在图片中框选目标区域,并标注框选类别。
3.如权利要求1所述一种基于图片识别的活体禽畜抵押贷款贷前估值方法,其特征在于,所述步骤S4中图片数据集按照8:2的比例随机划分训练集和测试集。
4.如权利要求1所述一种基于图片识别的活体禽畜抵押贷款贷前估值方法,其特征在于,所述步骤S5中离线训练yolov4动物数量统计模型利用卷积神经网络(CNN),同时提取适用于大多数模型、任务和数据集的通用特征。
5. 如权利要求4所述一种基于图片识别的活体禽畜抵押贷款贷前估值方法,其特征在于,所述通用特征的处理包括:使用加权残差连接(WRC)、阶段部分连接(CSP)、交叉小批量标准化(CmB N)、自对抗训练(SAT)、Mish激活、马赛克数据增强、CmBN、DropBlock正则化和CIoU丢失。
CN202011091917.9A 2020-10-13 2020-10-13 一种基于图片识别的活体禽畜抵押贷款贷前估值方法 Pending CN112184322A (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011091917.9A CN112184322A (zh) 2020-10-13 2020-10-13 一种基于图片识别的活体禽畜抵押贷款贷前估值方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011091917.9A CN112184322A (zh) 2020-10-13 2020-10-13 一种基于图片识别的活体禽畜抵押贷款贷前估值方法

Publications (1)

Publication Number Publication Date
CN112184322A true CN112184322A (zh) 2021-01-05

Family

ID=73949550

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011091917.9A Pending CN112184322A (zh) 2020-10-13 2020-10-13 一种基于图片识别的活体禽畜抵押贷款贷前估值方法

Country Status (1)

Country Link
CN (1) CN112184322A (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114781775A (zh) * 2022-01-10 2022-07-22 上海皓卡网络技术有限公司 一种多环节远程图像及信号管控的智慧管理系统
CN115311831A (zh) * 2022-08-01 2022-11-08 太湖流域管理局苏州管理局 一种电气柜监测系统及方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050257748A1 (en) * 2002-08-02 2005-11-24 Kriesel Marshall S Apparatus and methods for the volumetric and dimensional measurement of livestock
CN108805137A (zh) * 2018-04-17 2018-11-13 平安科技(深圳)有限公司 牲畜特征向量的提取方法、装置、计算机设备和存储介质
CN111709372A (zh) * 2020-06-18 2020-09-25 深圳市赛为智能股份有限公司 驱鸟方法、装置、计算机设备及存储介质
CN111709374A (zh) * 2020-06-18 2020-09-25 深圳市赛为智能股份有限公司 鸟情检测方法、装置、计算机设备及存储介质

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050257748A1 (en) * 2002-08-02 2005-11-24 Kriesel Marshall S Apparatus and methods for the volumetric and dimensional measurement of livestock
CN108805137A (zh) * 2018-04-17 2018-11-13 平安科技(深圳)有限公司 牲畜特征向量的提取方法、装置、计算机设备和存储介质
CN111709372A (zh) * 2020-06-18 2020-09-25 深圳市赛为智能股份有限公司 驱鸟方法、装置、计算机设备及存储介质
CN111709374A (zh) * 2020-06-18 2020-09-25 深圳市赛为智能股份有限公司 鸟情检测方法、装置、计算机设备及存储介质

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114781775A (zh) * 2022-01-10 2022-07-22 上海皓卡网络技术有限公司 一种多环节远程图像及信号管控的智慧管理系统
CN115311831A (zh) * 2022-08-01 2022-11-08 太湖流域管理局苏州管理局 一种电气柜监测系统及方法

Similar Documents

Publication Publication Date Title
CN107229947B (zh) 一种基于动物识别的金融保险方法及系统
CN112184322A (zh) 一种基于图片识别的活体禽畜抵押贷款贷前估值方法
US20220076068A1 (en) Method for constructing a convolution neural network based on farmland images, electronic device using the same
Lampe et al. Accounting for the wealth of Denmark: A case study of Smithian growth using the emergence of modern accounting in Danish dairying
Dutta A deep learning approach for animal breed classification-sheep
Mulinga Economic analysis of factors affecting technical efficiency of smallholders maize production on Rwanda
Skevas et al. A generalized true random-effects model with spatially autocorrelated persistent and transient inefficiency
Antle et al. Using AgMIP regional integrated assessment methods to evaluate vulnerability, resilience and adaptive capacity for climate smart agricultural systems
Marinchenko Transformation of the service sector as part of the agribusiness digitalization
US20210042582A1 (en) System and method for monitoring health of crop to validate crop insurance claim
Khairo et al. A study of technical inefficiencies of maize farmers within and outside the new agricultural extension program in the Harari region of Ethiopia
KR20200057440A (ko) 양돈 성장 조건 플랫폼 기반 출하 예측 제공 시스템 및 그 방법
Ansah et al. The effect of credit constraints on crop yield: evidence from soybean farmers in Northern Region of Ghana
Martins et al. Production environment and managerial techniques in explaining productivity growth in Brazilian beef cattle production
Moran et al. A study of key profit drivers in the Victorian dairy industry
Potdar et al. BAIF experience in field data collection
Chu-Ongsakul Challenges and opportunities of smart farming implementation.
Zsuzsanna The effect of agricultural subsidies on efficiency and productivity of farms: a literature review
Starkey Rapid appraisal methodologies for animal traction
Rezensky et al. the Role of Ownership Concentration As Moderation on Biological Asset Intensity, Company Size, and Firm Growth on Biological Asset Disclosure in Agricultural Industry in Indonesia
CN116343257A (zh) 基于深度学习的作种牲畜筛选方法、系统、设备及介质
Vigors Data collection methods used in the Beef Data and Genomics Programme (BDGP) and the development of restful API's for recording herd data
Puerto Agriculture, innovation, and development: What happens when new technology is not good enough?
Adebowale ENHANCING AGRICULTURAL GROWTH: EVALUATING THE IMPACT OF THE GROWTH ENHANCEMENT SUPPORT SCHEME (GESS) ON FOOD CROP FARMERS IN ANAMBRA STATE, NIGERIA
Evenson The Pretechnology Agricultural Sciences

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination