兽团平台犬鼻纹数据

当前确认犬只身份唯一性的方法为体内芯片植入式管理法，此方法目前因为操作难度大并未得到快捷有效的推广。十方物联研究犬只鼻纹识别法可以有效的推进犬只身份的录入登记，更快的将犬只数据数字化，为监管部门靶向监管、宠物保险、宠物医院及宠主生活提供技术支撑。数据来源：公司收集客户在平台处登记的宠物信息，并采用鼻纹识别方法将宠物信息录入数据库。
算法规则：兽团对鼻纹图像进行三方面处理：图像预处理、特征提取和特征匹配。图像预处理包括四个步骤：图像分割、滤波增强、二值化、细化，对鼻纹图像进行预处理后，去除了原图像的冗余部分，方便后续的识别处理；特征提取主要是提取鼻纹图像细化后的端点和分叉点；特征匹配是利用两个鼻纹的图像进行特征点比较，来确定两幅图像是否来自于同一犬只。算法给出了鼻纹图像预处理、特征提取、特征匹配的matlab程序及处理结果。该结果证明，用matlab实现的这些算法的处理结果比较理想，满足识别的可行性和应用性。
数据应用：将宠物信息录入系统后，可根据客户的要求匹配宠物信息，如宠物相亲、宠物走失找回等。

The current method for confirming the unique identity of dogs is the in-vivo chip implantation management method. However, this method has not been widely and efficiently promoted due to its high operational difficulty.

Shifang IoT's research on canine nose print recognition technology can effectively promote the entry and registration of dog identity information, accelerate the digitization of dog-related data, and provide technical support for targeted supervision by regulatory authorities, pet insurance, pet hospitals and the daily lives of pet owners.

Data Source: The company collects pet information registered by customers on its platform, and uses the nose print recognition method to input pet information into the database.

Algorithm Rules: The research team conducts three core processing steps on nose print images: image preprocessing, feature extraction and feature matching. Image preprocessing includes four steps: image segmentation, filtering enhancement, binarization and thinning. After preprocessing the nose print images, the redundant parts of the original images are removed, which facilitates subsequent recognition processing. Feature extraction mainly extracts the endpoints and bifurcation points of the thinned nose print images. Feature matching compares feature points from two nose print images to determine whether the two images come from the same dog. The algorithm provides the MATLAB programs and processing results for nose print image preprocessing, feature extraction and feature matching. The results prove that the processing effects of these algorithms implemented via MATLAB are satisfactory, meeting the requirements of recognition feasibility and applicability.

Data Application: After pet information is entered into the system, pet information matching can be performed according to customer requirements, such as pet dating and lost pet retrieval.

搜集汇总

数据集介绍

特点

该数据集包含106条犬只鼻纹及相关信息，用于通过鼻纹识别技术确认犬只身份唯一性，支持宠物保险、医院及监管等应用场景。