WO2020030054A1

WO2020030054A1 - Animal identification method and apparatus, medium and electronic device

Info

Publication number: WO2020030054A1
Application number: PCT/CN2019/099823
Authority: WO
Inventors: 罗扬
Original assignee: 京东数字科技控股有限公司
Priority date: 2018-08-10
Filing date: 2019-08-08
Publication date: 2020-02-13
Also published as: CN110826371A

Abstract

The embodiments of the present disclosure provide an animal identification method and apparatus, a medium and an electronic device, the method comprising: performing validity verification on acquired images, and filtering out an invalid image to obtain a first image; identifying the animal face in the first image, and marking the identified animal face to obtain a second image with the animal face marked therein; performing animal face feature identification on the second image, and marking the identified animal face features to form a third image, and generating a corresponding feature vector based on the identified animal face features; and achieving animal identification based on the third image and the feature vector corresponding to the third image. By means of the technical solution of the embodiments of the present disclosure, the costs are greatly saved on, the breeding efficiency is improved, and the entire breeding process is simplified and optimized, thereby providing basic support for subsequent unmanned and intelligent breeding, and improving the feasibility of implementing an artificial intelligence project in the animal husbandry industry. (FIG. 1)

Description

一种动物识别方法、装置、介质及电子设备Animal identification method, device, medium and electronic equipment

技术领域Technical field

本公开涉及基于图像的动物识别技术领域，具体而言，涉及一种动物识别方法及一种动物识别装置。The present disclosure relates to the technical field of image-based animal recognition, and in particular, to an animal recognition method and an animal recognition device.

背景技术Background technique

目前，我国的养猪业正在从传统的养猪业向现代化养猪业转变，但是，现有的猪场管理仍较为粗犷，从猪场建设到后期的饲养管理都缺乏技术人员的参与。由于市场行情因素的变化，诸多小散养殖户抗风险能力极差，无法保证其稳定盈利，而在饲养过程中，因消毒措施以及防范措施不到位，饲养人员与猪只的频繁接触的过程中造成细菌、疾病的传染亦是一大隐患。At present, the pig industry in our country is changing from the traditional pig industry to the modern pig industry. However, the existing pig farm management is still relatively rough. From the construction of the pig farm to the later rearing management, there is a lack of participation of technical personnel. Due to changes in market conditions, many small retail farmers have very poor anti-risk capabilities and cannot guarantee their stable profitability. During the breeding process, due to inadequate disinfection measures and preventive measures, the frequent contact between the breeder and the pigs Contagion of bacteria and diseases is also a major hidden danger.

目前养猪场在进行猪个体识别的常用方法有如下几种：There are several common methods for identifying individual pigs in pig farms:

(1)耳缺：一般在仔猪出生后1-2天内，根据相应的规则在猪耳的边缘，用耳缺钳剪出缺口，根据相应的规则组成数字编号，以识别不同的猪只。同一个猪场内，同一年份，同一品种猪的编号不可重复。此方法在行业内使用多年，是较为传统的编号方法。(1) Ear deficiencies: Generally within 1-2 days after piglets are born, cut out the gap with ear deficient forceps at the edge of the pig's ears according to the corresponding rules, and compose numbers according to the corresponding rules to identify different pigs. In the same pig farm, the same year, the number of the same breed of pigs must not be repeated. This method has been used in the industry for many years and is a more traditional numbering method.

(2)刺青：利用刺青钳对猪打上刺青，以分辨识别猪个体。(2) Tattoos: Tattoo pigs with tattoo forceps to identify individual pigs.

(3)耳标：使用时耳标头穿透牲畜耳部、嵌入辅标、固定耳标，耳标颈留在穿孔内。耳标面登载编码信息。多数情况下，对于留种后的成年种猪使用耳标，但现在也逐步对仔猪开始使用耳标。(3) Ear tag: When used, the ear tag penetrates the ear of the livestock, embeds the auxiliary tag, and fixes the ear tag. The ear tag neck is left in the perforation. The ear tag surface contains encoded information. In most cases, ear tags are used for adult breeding pigs after breeding, but ear tags are now gradually being used for piglets.

(4)标记：在进行接种疫苗以及需要识别猪只等活动中，通常会使用记号蜡笔进行标记。(4) Marking: Marking crayons are usually used for vaccination and pig identification.

而上述现有技术方案存在以下缺点：However, the foregoing prior art solutions have the following disadvantages:

(1)耳缺：不同的猪场在使用不同的打标标准，规范并不统一。有一些数字会出现错打现象，且无法纠正，读取过程中的错误率也较高，整个工作过程中产生的工作量十分巨大，在打标的过程中，对猪体本身亦有所伤害。(1) Ear deficiency: Different pig farms use different marking standards, and the specifications are not uniform. Some numbers will be mistyped and cannot be corrected. The error rate in the reading process is also high. The workload generated during the entire work process is huge. During the marking process, the pig body is also harmed. .

(2)刺青：在国内使用较少，操作过程较为繁琐，成本较高。(2) Tattoos: Less used in China, the operation process is more complicated, and the cost is higher.

(3)耳标：不同的猪需要不同规格的耳标，且在猪只活动时会导致耳标掉落，致使个体混淆，打标过程中产生大量人工成本。(3) Ear tags: Different pigs need ear tags of different specifications, and the ear tags will fall when the pigs are moving, resulting in confusion among individuals and a lot of labor costs during the marking process.

(4)标记：在利用记号蜡笔对猪只进行标记的过程中，可能会出现猪只乱跑的现象，导致标记困难，并出现重标，少标现象，在标记后一段时间内，猪只身上的标记就会慢慢褪去，又会出现猪只混淆的现象。(4) Marking: In the process of marking pigs with a marker crayon, the phenomenon of pigs running randomly may occur, resulting in difficulty in marking, and re-marking and under-marking phenomenon. Within a period of time after marking, pigs The marks on the body will slowly fade away, and pigs will be confused again.

需要说明的是，在上述背景技术部分公开的信息仅用于加强对本公开的背景的理解，因此可以包括不构成对本领域普通技术人员已知的现有技术的信息。It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to those of ordinary skill in the art.

发明内容Summary of the invention

本公开实施例的目的在于提供一种动物识别方法及一种动物识别装置，进而至少在一定程度上克服由于相关技术的限制和缺陷而导致的传统饲养动物识别需要耗费大量人工成本以及饲养动物标识容易出错等一个或者多个问题。The purpose of the embodiments of the present disclosure is to provide an animal identification method and an animal identification device, thereby at least to a certain extent overcoming the traditional breeding animal identification due to the limitations and defects of the related technology, which requires a lot of labor costs and raising animal identification Prone to errors and other problems.

本公开实施例的其他特性和优点将通过下面的详细描述变得显然，或部分地通过本公开的实践而习得。Other features and advantages of the embodiments of the present disclosure will become apparent from the following detailed description, or may be learned in part through the practice of the present disclosure.

根据本公开实施例的第一方面，提供一种动物识别方法，包括：According to a first aspect of the embodiments of the present disclosure, an animal identification method is provided, including:

对所获取的图像进行有效性校验，将无效图像滤除，获得第一图像；Perform validity check on the acquired image, filter out invalid images, and obtain a first image;

对所述第一图像中的动物面部进行识别，将所识别出的动物面部进行标识，获得标识出动物面部的第二图像；Identifying the animal face in the first image, identifying the identified animal face, and obtaining a second image identifying the animal face;

对所述第二图像进行动物面部特征识别，将所识别出的动物面部特征进行标注，形成第三图像，并基于所识别出的动物面部特征生成对应的特征向量；Performing animal facial feature recognition on the second image, labeling the identified animal facial features, forming a third image, and generating a corresponding feature vector based on the identified animal facial features;

基于所述第三图像和所述第三图像对应的特征向量实现动物识别。Animal recognition is achieved based on the third image and a feature vector corresponding to the third image.

在本公开的一个实施例中，上述获取的图像至少包括所述图像的数据格式信息、分辨率信息和动物面部；所述对所获取的图像进行有效性校验，将无效图像滤除，获得第一图像包括：In an embodiment of the present disclosure, the acquired image includes at least data format information, resolution information, and animal face of the image; the validity check is performed on the acquired image, and invalid images are filtered to obtain The first image includes:

将所述图像输入预设的有效性校验模型，获得所述图像的有效性校验结果；Inputting the image into a preset validity check model to obtain a validity check result of the image;

基于所述有效性校验结果，将不符合预设数据格式和预设分辨率的图像滤除，获得过滤后的图像；Filtering out images that do not conform to a preset data format and a preset resolution based on the validity check result to obtain a filtered image;

将所述过滤后的图像中的动物面部不完整的图像滤除，获得第一图像。The incomplete image of the animal face in the filtered image is filtered to obtain a first image.

在本公开的一个实施例中，上述方法还包括：In an embodiment of the present disclosure, the method further includes:

当判断所述图像不符合预设数据格式和预设分辨率，和\或判断所述图像中动物面部不完整时，生成并发送重新采集图像的控制信息。When it is determined that the image does not conform to a preset data format and a preset resolution, and / or it is determined that an animal face in the image is incomplete, control information for reacquiring the image is generated and sent.

在本公开的一个实施例中，上述第一图像中的动物面部进行识别，将所识别出的动物面部进行标识，获得标识出动物面部的第二图像，包括：In an embodiment of the present disclosure, identifying the animal face in the first image, identifying the identified animal face, and obtaining a second image identifying the animal face includes:

将所述第一图像输入至预设的动物面部识别模型，获得与所述第一图像对应的动物面部识别结果；Inputting the first image to a preset animal face recognition model to obtain an animal face recognition result corresponding to the first image;

基于所述动物面部识别结果，确定所述第一图像中动物面部范围；Determining an animal face range in the first image based on the animal face recognition result;

将所述动物面部范围框画在方框内，生成框画出动物面部范围的第二图像。Draw the frame of the animal face within the frame, and generate a second image that draws the frame of the animal face.

在本公开的一个实施例中，上述第二图像进行动物面部特征识别，将所识别出的动物面部特征进行标注，形成第三图像，包括：In an embodiment of the present disclosure, the second image performs animal facial feature recognition, and the identified animal facial features are labeled to form a third image, including:

识别出所述第二图像动物面部的各部位；Identifying parts of the animal face in the second image;

按照预设的顺序对所识别出的各个部位进行标注，获得带有动物面部的各个部位标注的第三图像。Annotate each identified part in a preset order to obtain a third image with annotated parts of the animal's face.

在本公开的一个实施例中，上述各部位至少包括：眼睛部位、鼻子部位、耳朵部位；所述按照预设的顺序对所识别出的各个部位进行标注，获得带有动物面部的各个部位标注的第三图像，包括：In an embodiment of the present disclosure, each of the above parts includes at least: an eye part, a nose part, and an ear part; the identified parts are labeled in a preset order, and the parts with animal faces are labeled. The third image includes:

在所述第二图像中所识别出眼睛部位的左眼和右眼标注出位置点；The left eye and the right eye of the eye part identified in the second image are marked with position points;

在所述第二图像中所识别出鼻子部位的左鼻尖和右鼻尖标注出位置点；The left nose tip and the right nose tip of the nose part identified in the second image are marked with position points;

在所述第二图像中所识别出耳朵部位的左耳根部和右耳根部标注出位置点；Mark the location points of the left ear root and the right ear root of the ear part identified in the second image;

获得有标注出所述眼睛部位左眼和右眼的位置点、所述鼻子部位左鼻尖和右鼻尖的位置点、所述耳朵部位左耳根部和右耳根的位置点的第三图像。A third image is obtained with the positions of the left and right eyes of the eye part, the positions of the left and right nose tip of the nose part, and the positions of the left and right ear base of the ear part.

在本公开的一个实施例中，上述基于所识别出的动物面部特征生成对应的特征向量，包括：In an embodiment of the present disclosure, the above-mentioned generating a corresponding feature vector based on the identified animal facial features includes:

根据所述第三图像中动物面部的眼睛部位左眼和右眼的位置点、所述鼻子部位左鼻尖和右鼻尖的位置点、所述耳朵部位左耳根部和右耳根的位置点与所述框画出动物面部范围方框的相对位置，生成所述动物面部的特征向量。According to the position points of the left and right eyes of the eye part of the animal face in the third image, the position points of the left and right nose tips of the nose part, and the position points of the left and right ear roots of the ear part and the The frame draws the relative position of the animal face range box, and generates a feature vector of the animal face.

根据本公开实施例的第二方面，提供一种动物识别装置，包括：校验模块、识别模块、特征识别模块、输出模块；其中，According to a second aspect of the embodiments of the present disclosure, an animal identification device is provided, including: a verification module, an identification module, a feature identification module, and an output module; wherein,

校验模块，用于对所获取的图像进行有效性校验，将无效图像滤除，获得第一图像；A verification module, configured to verify the validity of the acquired image, filter out invalid images, and obtain a first image;

识别模块，用于对所述第一图像中的动物面部进行识别，将所识别出的动物面部进行标识，获得标识出动物面部的第二图像；A recognition module, configured to recognize an animal face in the first image, identify the identified animal face, and obtain a second image identifying the animal face;

特征识别模块，用于对所述第二图像进行动物面部特征识别，将所识别出的动物面部特征进行标注，形成第三图像，并基于所识别出的动物面部特征生成对应的特征向量；A feature recognition module, configured to perform animal facial feature recognition on the second image, mark the identified animal facial features, form a third image, and generate a corresponding feature vector based on the identified animal facial features;

输出模块，用于基于所述第三图像和所述第三图像对应的特征向量实现动物识别。An output module is configured to implement animal recognition based on the third image and a feature vector corresponding to the third image.

根据本公开实施例的第三方面，提供了一种计算机可读介质，其上存储有计算机程序，所述程序被处理器执行时实现如上述实施例中第一方面所述的动物识别方法。According to a third aspect of the embodiments of the present disclosure, there is provided a computer-readable medium having stored thereon a computer program that, when executed by a processor, implements the animal identification method as described in the first aspect of the above embodiments.

根据本公开实施例的第四方面，提供了一种电子设备，包括：一个或多个处理器；存储装置，用于存储一个或多个程序，当所述一个或多个程序被所述一个或多个处理器执行时，使得所述一个或多个处理器实现如上述实施例中第一方面所述的动物识别方法。According to a fourth aspect of the embodiments of the present disclosure, there is provided an electronic device including: one or more processors; a storage device for storing one or more programs, and when the one or more programs are used by the one When executed by one or more processors, the one or more processors are caused to implement the animal identification method according to the first aspect in the foregoing embodiment.

本公开实施例提供的技术方案可以包括以下有益效果：The technical solutions provided by the embodiments of the present disclosure may include the following beneficial effects:

本公开的一些实施例所提供的技术方案中，对所获取的图像进行有效性校验，将无效图像滤除，获得第一图像；对第一图像中的动物面部进行识别，将所识别出的动物面部进行标识，获得标识出动物面部的第二图像；对第二图像进行动物面部特征识别，将所识别出的动物面部特征进行标注，形成第三图像，并基于所识别出的动物面部特征生成对应的特征向量；基于第三图像和第三图像对应的特征向量实现动物识别。本公开实施例的技术方案通过人工智能算法建立动物的特征信息库，实现对动物进行建档或识别，极大节约了成本，提高了饲养效率，简化并优化了整个饲养流程，为后续的无人化、智能化饲养提供基础支撑，提升了人工智能项目在畜牧行业落地的可行性。In the technical solutions provided by some embodiments of the present disclosure, a validity check is performed on the acquired image, and the invalid image is filtered to obtain a first image; the animal face in the first image is recognized, and the identified Identify the animal's face to obtain a second image identifying the animal's face; identify the animal's facial features on the second image; mark the identified animal's facial features to form a third image; and based on the identified animal's face The feature generates a corresponding feature vector; animal recognition is achieved based on the third image and the feature vector corresponding to the third image. The technical solution of the embodiment of the present disclosure establishes an animal characteristic information database through an artificial intelligence algorithm, and realizes archiving or identification of animals, which greatly saves costs, improves breeding efficiency, and simplifies and optimizes the entire breeding process. Humanized and intelligent feeding provides basic support, which improves the feasibility of artificial intelligence projects in the livestock industry.

应当理解的是，以上的一般描述和后文的细节描述仅是示例性和解释性的，并不能限制本公开。It should be understood that the above general description and the following detailed description are merely exemplary and explanatory, and should not limit the present disclosure.

附图说明BRIEF DESCRIPTION OF THE DRAWINGS

此处的附图被并入说明书中并构成本说明书的一部分，示出了符合本公开的实施例，并与说明书一起用于解释本公开的原理。显而易见地，下面描述中的附图仅仅是本公开的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。在附图中：The drawings herein are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure, and together with the description serve to explain the principles of the present disclosure. Obviously, the drawings in the following description are just some embodiments of the present disclosure. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts. In the drawings:

图1示意性示出了根据本公开的一个实施例的动物识别方法的流程图。FIG. 1 schematically illustrates a flowchart of an animal identification method according to an embodiment of the present disclosure.

图2A示意性示出了根据本公开的一个实施例的图像存储于云端的示意图；FIG. 2A is a schematic diagram illustrating that an image is stored in the cloud according to an embodiment of the present disclosure; FIG.

图2B示意性示出了根据本公开的一个实施例的图像存储于本地的示意图；FIG. 2B schematically illustrates that an image is stored locally according to an embodiment of the present disclosure; FIG.

图3示意性示出了根据本公开的一个实施例的框画出动物面部范围图像的示意图；FIG. 3 schematically illustrates a frame drawing an animal face range image according to an embodiment of the present disclosure; FIG.

图4示意性示出了根据本公开的一个实施例的对第二图像中动物面部特征进行标注后的示意图；4 schematically illustrates a schematic diagram after labeling animal facial features in a second image according to an embodiment of the present disclosure;

图5示意性示出了根据本公开的一个实施例动物识别方法应用于猪只个体识别的流程图；5 schematically illustrates a flowchart of an animal identification method applied to individual pig identification according to an embodiment of the present disclosure;

图6示意性示出了根据本公开的一个实施例的动物识别装置的框图；FIG. 6 schematically illustrates a block diagram of an animal identification device according to an embodiment of the present disclosure;

图7示出了适于用来实现本公开实施例的电子设备的计算机***的结构示意图。FIG. 7 shows a schematic structural diagram of a computer system suitable for implementing an electronic device according to an embodiment of the present disclosure.

具体实施方式detailed description

现在将参考附图更全面地描述示例实施方式。然而，示例实施方式能够以多种形式实施，且不应被理解为限于在此阐述的范例；相反，提供这些实施方式使得本公开将更加全面和完整，并将示例实施方式的构思全面地传达给本领域的技术人员。Example embodiments will now be described more fully with reference to the accompanying drawings. However, the example embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, the embodiments are provided so that this disclosure will be more comprehensive and complete, and the concepts of the example embodiments will be fully conveyed To those skilled in the art.

此外，所描述的特征、结构或特性可以以任何合适的方式结合在一个或更多实施例中。在下面的描述中，提供许多具体细节从而给出对本公开的实施例的充分理解。然而，本领域技术人员将意识到，可以实践本公开的技术方案而没有特定细节中的一个或更多，或者可以采用其它的方法、组元、装置、步骤等。在其它情况下，不详细示出或描述公知方法、装置、实现或者操作以避免模糊本公开的各方面。Furthermore, the described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, many specific details are provided to give a full understanding of the embodiments of the present disclosure. However, those skilled in the art will realize that the technical solutions of the present disclosure may be practiced without one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other instances, well-known methods, devices, implementations or operations have not been shown or described in detail to avoid obscuring aspects of the present disclosure.

附图中所示的方框图仅仅是功能实体，不一定必须与物理上独立的实体相对应。即，可以采用软件形式来实现这些功能实体，或在一个或多个硬件模块或集成电路中实现这些功能实体，或在不同网络和/或处理器装置和/或微控制器装置中实现这些功能实体。The block diagrams shown in the drawings are merely functional entities and do not necessarily correspond to physically separate entities. That is, these functional entities may be implemented in the form of software, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices. entity.

附图中所示的流程图仅是示例性说明，不是必须包括所有的内容和操作/步骤，也不是必须按所描述的顺序执行。例如，有的操作/步骤还可以分解，而有的操作/步骤可以合并或部分合并，因此实际执行的顺序有可能根据实际情况改变。The flowcharts shown in the accompanying drawings are only exemplary descriptions, and it is not necessary to include all contents and operations / steps, nor are they necessarily performed in the order described. For example, some operations / steps can also be decomposed, and some operations / steps can be merged or partially merged, so the order of actual execution may be changed according to the actual situation.

参照图1所示，根据本公开的一个实施例的动物识别方法，包括以下步骤：Referring to FIG. 1, an animal identification method according to an embodiment of the present disclosure includes the following steps:

步骤S110，对所获取的图像进行有效性校验，将无效图像滤除，获得第一图像；Step S110: Perform validity check on the acquired image, filter out invalid images, and obtain a first image;

步骤S120，对第一图像中的动物面部进行识别，将所识别出的动物面部进行标识，获得标识出动物面部的第二图像；Step S120: identifying the animal face in the first image, identifying the identified animal face, and obtaining a second image identifying the animal face;

步骤S130，对第二图像进行动物面部特征识别，将所识别出的动物面部特征进行标注，形成第三图像，并基于所识别出的动物面部特征生成对应的特征向量；Step S130: Perform animal facial feature recognition on the second image, mark the identified animal facial features, form a third image, and generate a corresponding feature vector based on the identified animal facial features;

步骤S140，当基于第三图像和第三图像对应的特征向量实现动物识别。In step S140, animal recognition is implemented based on the third image and the feature vector corresponding to the third image.

图1所示实施例的技术方案能够通过人工智能算法建立动物的特征信息库，实现对动物进行建档或识别，极大节约了成本，提高了饲养效率，简化并优化了整个饲养流程，为后续的无人化、智能化饲养提供基础支撑，提升了人工智能项目在畜牧行业落地的可行性。The technical solution of the embodiment shown in FIG. 1 can establish an animal characteristic information database through an artificial intelligence algorithm, and realize the archiving or identification of animals, which greatly saves costs, improves breeding efficiency, and simplifies and optimizes the entire breeding process. Subsequent unmanned and intelligent feeding provides basic support, which improves the feasibility of artificial intelligence projects in the livestock industry.

以下对图1中所示的各个步骤的实现细节进行详细阐述：The implementation details of each step shown in FIG. 1 are described in detail below:

在步骤S110中，对所获取的图像进行有效性校验，将无效图像滤除，获得第一图像。In step S110, a validity check is performed on the acquired image, and invalid images are filtered to obtain a first image.

在本公开的一个实施例中，上述步骤S110之前，还包括：通过预先架设的摄录设备获取图像，具体的，预先架设的摄录设备可以是饲养环境已架设的用于监控饲养动物整体运行情况的摄录设备，例如：摄像头、相机、手机等，通过该摄录设备也可以为人工智能动物识别算法提供相应的图像。In an embodiment of the present disclosure, before step S110, the method further includes: obtaining an image through a pre-set camera and recording device. Specifically, the pre-set camera and recording device may be set up in a breeding environment for monitoring the overall operation of the breeding animal. The recording device of the situation, such as a camera, a camera, a mobile phone, etc., can also provide corresponding images for the artificial intelligence animal recognition algorithm through the recording device.

在本公开的一个实施例中，基于前述方案，上述摄录设备的架设可以根据实际需求进行调整，根据识别所需求的图像质量、视频质量、光线等原因，选取相应参数的摄录设备进行架设。In an embodiment of the present disclosure, based on the foregoing solution, the installation of the above-mentioned recording equipment can be adjusted according to actual needs, and according to the requirements of identifying the required image quality, video quality, light, etc., the corresponding parameters are selected for installation. .

在本公开的一个实施例中，基于前述方案，步骤S110是对所获取的图像进行有效性检测以及识别动物面部是否完整，因此，在步骤S110之前架设摄录设备的过程中需要从多个角度对动物进行拍摄，以在后续的识别过程中尽可能从多个角度拍摄到动物的面部，以能够在数据库中对动物建立多个面部数据，为后期能够准确的识别出动物提供数据基础。In an embodiment of the present disclosure, based on the foregoing solution, step S110 is to perform validity detection on the acquired image and identify whether the animal's face is complete. Therefore, during the process of setting up the recording device before step S110, multiple angles need to be set. The animals are photographed to capture the animal's face from multiple angles as much as possible in the subsequent recognition process, so as to be able to establish multiple facial data for the animal in the database, and provide a data basis for accurate animal identification in the later stage.

在本公开的一个实施例中，基于前述方案，可以根据饲养动物环境的不同，以及需要在夜间或不同光照条件下进行动物拍摄的情况下，在不同的位置选用不同参数的摄录设备，以降低整体的硬件成本。In one embodiment of the present disclosure, based on the foregoing scheme, according to different animal breeding environments, and in the case that animals need to be photographed at night or under different lighting conditions, camera equipment with different parameters is selected at different locations to Reduce overall hardware costs.

在本公开的一个实施例中，基于前述方案，上述根据动物饲养环境所铺设的摄录设备是提供图像的基础设施，可以根据不同的存储方式将摄录设备所摄录视频进行保存，参照图2a所示，所拍摄的视频可以存储在云端，或参照图2b所示，存储在本地外接的设备存储中，在后期通过实时视频流的传输，对视频进行抽取关键帧，根据抽出的关键帧对摄录范围内的动物进行检测与识别。In an embodiment of the present disclosure, based on the foregoing solution, the above-mentioned video recording device laid according to the animal breeding environment is an infrastructure for providing images, and the video recorded by the video recording device can be saved according to different storage methods. As shown in 2a, the captured video can be stored in the cloud, or as shown in Figure 2b, stored in a local external device storage. In the later stage, the video is extracted through key frames, and the key frames are extracted according to the extracted key frames. Detection and identification of animals in the recording range.

在本公开的一个实施例中，动物面部检测与识别的产出是基于算法对图像进行加工后的数据，将采集的原始图像通过接口或者批量导入的形式录入到***中，等待识别装置的处理，如果摄像头选择的是云端存储，则需要从第三方接入视频流接口，以获取实时图像或视频数据；如果摄像头选择的是本地外接存储设备存储，则需要有本地服务器或者通过网络实时回传图像或视频数据。In an embodiment of the present disclosure, the output of animal face detection and recognition is based on the data processed by the algorithm based on the algorithm. The collected original images are entered into the system through the interface or batch import, waiting for processing by the recognition device. If the camera chooses cloud storage, you need to access the video streaming interface from a third party to get real-time images or video data; if the camera chooses a local external storage device to store, you need a local server or real-time backhaul through the network Image or video data.

在本公开的一个实施例中，所获取的图像至少包括图像的数据格式信息、分辨率信息和动物面部，上述步骤S110中具体包括：将图像输入预设的有效性校验模型，获得图像的有效性校验结果；基于有效性校验结果，将不符合预设数据格式和预设分辨率的图像滤除，获得过滤后的图像；将过滤后的图像中的动物面部不完整的图像滤除，获得第一图像。In an embodiment of the present disclosure, the acquired image includes at least data format information, resolution information, and animal face of the image. The above step S110 specifically includes: inputting the image into a preset validity check model to obtain the image. Validity check results; based on the validity check results, filtering out images that do not meet the preset data format and preset resolution to obtain filtered images; filtering incomplete images of animal faces in the filtered images Divide to obtain the first image.

在本公开的一个实施例中，基于前述方案，当判断图像不符合预设数据格式和预设分辨率，和\或判断图像中动物面部不完整时，生成并发送重新采集图像的控制信息。In an embodiment of the present disclosure, based on the foregoing scheme, when it is determined that the image does not conform to the preset data format and preset resolution, and / or it is determined that the animal face in the image is incomplete, control information for reacquiring the image is generated and sent.

在本公开的一个实施例中，当从摄录设备拍摄的视频中提取到图像后，对该图像的真实性和有效性进行校验，可以根据动物面部的识别规范或条件，形成的简单的有效性校验算法，首先，对输入的图像进行一次初步的过滤和筛查，过滤图像格式和分辨率不符合要求的图像，最后，从过滤后的图像中判断动物面部是否有截断，将动物面部不完整的图像滤除，获得符合图像格式和分辨率要求以及动物面部完整的第一图像；In an embodiment of the present disclosure, after an image is extracted from a video captured by a video recording device, the authenticity and validity of the image can be verified, which can be formed according to animal face recognition specifications or conditions. Validity check algorithm. First, perform a preliminary filtering and screening on the input image, filter the images whose image format and resolution do not meet the requirements, and finally, determine whether the animal face is truncated from the filtered image, Filter out incomplete face images to obtain the first image that meets the image format and resolution requirements and the complete face of the animal;

在本公开的一个实施例中，基于前述方案，将图像格式和分辨率不符合的图像以及动物面部不完成的图像定义为异常图像，直接退出动物面部识别流程，不再占用***资源进行处理。In one embodiment of the present disclosure, based on the foregoing scheme, images with inconsistent image formats and resolutions and incomplete images of animal faces are defined as abnormal images, and the animal face recognition process is directly exited without using system resources for processing.

在本公开的一个实施例中，在后续基于识别出的动物面部进行建档的流程中，由于动物建档过程中会采集大量数据，因此，通过上述对图像的有效性校验流程可以实现只根据有效图像进行建档，避免了根据无效图像建档而导致的无法准确对动物进行识别的问题；若在建档后对动物进行识别的使用过程中出现无效数据，则应返回数据采集部分，重新采集动物面部图像以对其再次进行识别。In an embodiment of the present disclosure, in the subsequent process of archiving based on the identified animal face, since a large amount of data is collected during the archival of the animal, the above-mentioned validity verification process for the image can be used to achieve only Archiving based on valid images avoids the problem of inaccurate identification of animals caused by archiving of invalid images; if invalid data appears during the process of identifying animals after archiving, you should return to the data collection section. The animal's face image is reacquired to recognize it again.

在步骤S120中，对第一图像中的动物面部进行识别，将所识别出的动物面部进行标识，获得标识出动物面部的第二图像。In step S120, the animal face in the first image is identified, the identified animal face is identified, and a second image identifying the animal face is obtained.

在本公开的一个实施例中，将第一图像输入至预设的动物面部识别模型，获得与第一图像对应的动物面部识别结果；基于动物面部识别结果，确定第一图像中动物面部范围；将动物面部范围框画在方框内，生成框画出动物面部范围的第二图像。In an embodiment of the present disclosure, a first image is input to a preset animal face recognition model to obtain an animal face recognition result corresponding to the first image; and an animal face range in the first image is determined based on the animal face recognition result; Draw the frame of the animal's face within the frame, and generate a second image of the frame of the animal's face.

在本公开的一个实施例中，完成步骤S110中图像的有效性判断后，对这第一图像进行人工智能算法检测，具体的，可以是对输入的数据进行动物面部检测，即为检测出图像中是否存在动物面部，若存在则将动物面部框出。In one embodiment of the present disclosure, after the validity judgment of the image in step S110 is completed, the first image is detected by an artificial intelligence algorithm. Specifically, the input data may be detected by an animal face, that is, the detected image Whether there is an animal face, and if so, frame the animal face.

图3示意性示出了根据本公开的一个实施例的框画出动物面部范围图像的示意图。FIG. 3 schematically illustrates a frame drawing an animal face range image according to an embodiment of the present disclosure.

如图3所示，在步骤S120中，需要对输入的第一图像进行处理，分析图像中的动物面部，对输入的第一图像中的动物面部进行画框处理，每一只动物的面部均需框出，为后期进行识别提供支持。As shown in FIG. 3, in step S120, it is necessary to process the input first image, analyze the animal face in the image, and perform frame processing on the animal face in the input first image. The face of each animal is It needs to be framed to provide support for later identification.

在步骤S130中，对第二图像进行动物面部特征识别，将所识别出的动物面部特征进行标注，形成第三图像，并基于所识别出的动物面部特征生成对应的特征向量。In step S130, animal facial feature recognition is performed on the second image, the identified animal facial features are labeled, a third image is formed, and a corresponding feature vector is generated based on the identified animal facial features.

在本公开的一个实施例中，识别出第二图像动物面部的各部位；按照预设的顺序对所识别出的各个部位进行标注，获得带有动物面部的各个部位标注的第三图像。In one embodiment of the present disclosure, each part of the animal face in the second image is identified; the identified parts are labeled in a preset order to obtain a third image with the parts of the animal face labeled.

在本公开的一个实施例中，基于前述方案，在第二图像中所识别出眼睛部位的左眼和右眼标注出位置点；在第二图像中所识别出鼻子部位的左鼻尖和右鼻尖标注出位置点；在第二图像中所识别出耳朵部位的左耳根部和右耳根部标注出位置点；获得有标注出眼睛部位左眼和右眼的位置点、鼻子部位左鼻尖和右鼻尖的位置点、耳朵部位左耳根部和右耳根的位置点的第三图像。In an embodiment of the present disclosure, based on the foregoing scheme, the left and right eyes of the eye part identified in the second image are marked with position points; and the left and right nose tips of the nose part are identified in the second image. Mark the location points; mark the location points of the left and right ear roots of the ear area identified in the second image; obtain the location points of the left and right eyes of the eye area, and the left and right nose points of the nose area A third image of the location point of the ear, the location of the left ear root and the right ear root of the ear.

图4示意性示出了根据本公开的一个实施例的对第二图像中动物面部特征进行标注后的示意图。FIG. 4 schematically illustrates a schematic diagram after labeling animal facial features in a second image according to an embodiment of the present disclosure.

如图4所示，在本公开的一个实施例中，将第二图像输入到动物面部识别算法模型中，由相应的动物面部特征识别算法对检测出的动物面部进行识别，识别出动物面部的各部位，在步骤S130中，对检测出的动物面部的各部位进行打点处理，按照一定的顺序在动物面部的各部位例如：左鼻尖、右鼻尖、左眼睛、右眼睛、左耳根部以及右耳根部进行打点，对上述各部位打点后，输出带有猪脸识别标记点的第三图像，同时输出一个动物面部的位置坐标向量，根据不同动物面部特征点标记位置的向量值的不同来对其进行识别，由于动物个体之间长相的差别，因此，每只动物均会输出一个唯一向量。As shown in FIG. 4, in an embodiment of the present disclosure, a second image is input into an animal face recognition algorithm model, and the detected animal face is recognized by a corresponding animal face feature recognition algorithm, and the animal face is identified. For each part, in step S130, dot detection is performed on each part of the detected animal face, and the parts of the animal face are, for example, left nose tip, right nose tip, left eye, right eye, left ear root, and right in a certain order. Dotting on the root of the ear, and outputting a third image with pig face recognition marks after marking the above parts, and simultaneously outputting an animal face position coordinate vector. It recognizes that each animal will output a unique vector due to the difference in looks between the individual animals.

在本公开的一个实施例中，动物识别可以是对动物进行建档，或对动物进行识别，其中，对动物进行建档包括：当输出标注有动物特征的第三图片和之对应的特征向量后，判断第三图片和之对应的特征向量是否为建档数据，若为建档数据，则应将第三图片和之对应的特征向量输入数据库中，在数据库中根据此数据建立动物档案，并为该动物按照一定的编号顺序编制唯一编号，为后续使用过程中提供支持；对动物进行识别包括：若判断第三图片和之对应的特征向量不是建档数据，则应将第三图片和之对应的特征向量与数据库中的动物的档案数据进行比对，寻找到该动物数据，并将其编号输出，同时调出其相应的管理数据，以便供使用者进行处理，对工作提供帮助。In an embodiment of the present disclosure, the animal identification may be to file an animal, or to identify the animal, wherein the animal filed includes: when outputting a third picture labeled with animal characteristics and a corresponding feature vector Then, it is judged whether the third picture and the corresponding feature vector are archival data. If it is the archival data, the third picture and the corresponding feature vector should be entered into the database, and an animal file is created based on this data in the database. A unique number is prepared for the animal in a certain numbering sequence to provide support for subsequent use. Identifying the animal includes: if the third picture and its corresponding feature vector are not archival data, the third picture and the The corresponding feature vector is compared with the file data of the animal in the database, the animal data is found, and its number is output. At the same time, the corresponding management data is called up for the user to process and help the work.

本公开实施例所提供的一种动物识别方法通过对所获取的图像进行有效性校验，将无效图像滤除，获得第一图像；对第一图像中的动物面部进行识别，将所识别出的动物面部进行标识，获得标识出动物面部的第二图像；对第二图像进行动物面部特征识别，将所识别出的动物面部特征进行标注，形成第三图像，并基于所识别出的动物面部特征生成对应的特征向量；基于第三图像和第三图像对应的特征向量实现动物识别。本公开实施例的技术方案通过人工智能算法建立动物的特征信息库，实现对动物进行建档或识别，解决了人工所饲养的动物无法识别，对不同动物实行针对性措施困难等现象，能够准确实时的掌握饲养动物的全部状态(如日龄、孕周、分娩次数、身体状态、是否接种疫苗等)，并且直观可视；针对不同动物在不同的时间节点对工作人员进行提醒，合理安排饲养员工作，节约大量的等待时间，为后续的智能化无人猪场提供基础支撑；所铺设摄录设备以最优布局进行设置，使得每个摄像头的利用率尽可能提高，根据不同位置的图像要求将设置不同参数的摄像头，在满足要求的前提下，尽可能降低成本；本公开实施例的技术方案为先检测后识别，且输出的特征向量值唯一，能够保证识别出动物，从动物的正脸到侧脸，多角度为动物建立档案，保证了档案数据的唯一性；通过优化设计的数据接入方案，能够在需要时随时截取有效数据，并根据不同饲养场的实际情况选用不同方案，实现个性化定制。An animal recognition method provided by an embodiment of the present disclosure performs a validity check on an acquired image, filters out invalid images to obtain a first image, and recognizes an animal face in the first image to identify the identified Identify the animal's face to obtain a second image identifying the animal's face; identify the animal's facial features on the second image; mark the identified animal's facial features to form a third image; and based on the identified animal's face The feature generates a corresponding feature vector; animal recognition is achieved based on the third image and the feature vector corresponding to the third image. The technical solution of the embodiment of the present disclosure establishes an animal characteristic information database through an artificial intelligence algorithm, and realizes archiving or identification of animals, and solves the problems that artificially-bred animals cannot be identified, and it is difficult to implement targeted measures for different animals. Real-time grasp of all the status of the animals (such as age, gestational week, number of deliveries, physical status, whether to be vaccinated, etc.), and visually visible; remind staff at different time nodes for different animals, and arrange rearing reasonably Staff, save a lot of waiting time, and provide basic support for subsequent intelligent unmanned pig farms; the installed video recording equipment is set in an optimal layout to make the utilization of each camera as high as possible, according to the images of different locations It is required that cameras with different parameters will be set to reduce costs as much as possible under the premise of meeting the requirements; the technical solution of the embodiment of the present disclosure is to detect first and then identify, and the output feature vector value is unique, which can ensure that animals are identified, Face-to-face, multi-angle file creation for animals, ensuring the number of files Uniqueness; by optimizing the design of the data access, can be taken at any time when valid data is required, and the choice of different programs according to the actual situation of the different farms achieve customization.

需要说明的是，上述内容，仅为本公开的较佳实施例而已，并非用于限定本公开的保护范围。It should be noted that the foregoing is merely a preferred embodiment of the present disclosure, and is not intended to limit the protection scope of the present disclosure.

以下介绍本公开所提出动物识别方法应用于识别猪只个体的实施例。The following describes an embodiment in which the animal identification method proposed in the present disclosure is applied to identify individual pigs.

图5示意性示出了根据本公开的一个实施例动物识别方法应用于猪只个体识别的流程图。FIG. 5 schematically illustrates a flowchart of an animal identification method applied to individual pig identification according to an embodiment of the present disclosure.

参照图5所示，根据本公开的一个实施例的动物识别方法应用于猪只个体识别的流程，包括以下步骤：Referring to FIG. 5, a process for applying an animal identification method according to an embodiment of the present disclosure to individual pig identification includes the following steps:

步骤S501，在猪饲养场架设摄录设备；Step S501, setting up a recording device at a pig breeding farm;

在本公开的一个实施例中，摄录设备的架设是为了监控猪场整体运行情况以便及时发现问题，同时也为人工智能算法提供相应的图片数据。根据猪只不同部位所需要图片、视频质量的不同，选取相应参数的摄录设备进行架设；本流程的主要功能是为了检测、识别猪脸，因此，在架设摄录设备的过程中需要有不同角度，在建档过程中需要有尽可能多的角度拍摄到一头猪的头部，建立尽可能多的猪只脸部数据，为后期能够更准确的识别提供数据基础；根据不同的场地情况以及建档需求，可以选择不同的摄录设备(例：摄像头，相机，手机、移动终端等)。根据不同实际情况，有时需要在夜间或者不同光照条件下进行摄录，在考虑成本的同时，需要在不同的位置选用不同参数的摄录设备，从整体降低硬件成本。In one embodiment of the present disclosure, the recording device is set up to monitor the overall operation of the pig farm in order to detect problems in a timely manner, and at the same time provide corresponding image data for the artificial intelligence algorithm. According to the different pictures and video quality of different parts of the pig, the camera with the corresponding parameters is selected for installation. The main function of this process is to detect and identify the pig's face. Therefore, the process of setting up the camera needs to be different. Angle, during the process of file creation, it is necessary to photograph the head of a pig with as many angles as possible, to establish as much pig face data as possible, to provide a data basis for more accurate identification in the later stage; according to different site conditions and For archival requirements, you can choose different recording and recording equipment (for example: camera, camera, mobile phone, mobile terminal, etc.). According to different actual situations, sometimes it is necessary to record at night or under different lighting conditions. While considering the cost, it is necessary to select different parameters of recording equipment at different locations to reduce the overall hardware cost.

步骤S502，通过所架设的摄录设备采集图像；Step S502, acquiring images through the set up camera equipment;

在本公开的一个实施例中，前期架设的摄录设备是提供图像的基础设施，架设的摄录设备所采集的图像可以存储在云端，或者在本地外接存储设备急性存储，在实际应用中，可以根据不同的存储方式将摄录设备所摄录视频进行保存，在后期通过实时视频流的传输，利用算法对视频进行抽帧处理，根据抽出的帧对摄录范围内的猪只进行检测与识别。同时要求摄像头能够随时拍摄到有效画面，能够保证画质清晰，并存储为后期算法可识别格式，在拍摄过程中要尽可能的保证每头猪脸有尽可能全的角度，以提高后期猪脸识别的准确率。In an embodiment of the present disclosure, the pre-recorded recording device is an infrastructure for providing images. The images collected by the set-up recording device can be stored in the cloud, or can be stored acutely in a local external storage device. In practical applications, The video recorded by the recording device can be saved according to different storage methods, and the real-time video stream is transmitted in the later stage. The video is framed by an algorithm, and the pigs in the recording range are detected and detected based on the extracted frames. Identify. At the same time, it is required that the camera can take effective pictures at any time, ensure the image quality is clear, and store it in a format that can be recognized by the post-processing algorithm. During the shooting process, it is necessary to ensure that each pig face has the fullest possible angle to improve the post-face Identification accuracy.

步骤S503，将所采集的图像录入；Step S503: input the acquired image;

在本公开的一个实施例中，猪脸检测与识别的产出是基于算法对信息进行加工后的数据。将采集的原始数据通过接口或者批量导入的形式录入到***中，等待人工智能***的处理，如果摄像头将采集的图像在云端进行存储，则需要从第三方接入视频流接口，以获取实时数据；如果摄像头将采集的图像在本地外接存储设备进行存储，则需要有本地服务器或者通过网络实时回传视频流。In one embodiment of the present disclosure, the output of pig face detection and recognition is data based on an algorithm that processes information. Enter the collected raw data into the system through the interface or batch import, and wait for processing by the artificial intelligence system. If the camera stores the collected images in the cloud, you need to access the video stream interface from a third party to obtain real-time data ; If the camera stores the captured images on a local external storage device, you need a local server or a real-time video stream back through the network.

在实际应用中可以根据不同的情况选取不同的存储方式，由于某些云端存储在上传云端的过程中，会对视频进行压缩，因此并不能够从云端得到原画质的视频，如果在后期对视频画面的要求较高的情况下，应选用能够满足其要求的存储方式，具体内容可参照图2a和图2b，这里不再赘述。In practical applications, different storage methods can be selected according to different situations. Since some cloud storage compresses the video during the upload to the cloud, it is not possible to get the original image quality video from the cloud. When the requirements of the video picture are high, a storage method capable of meeting the requirements should be selected. For details, refer to FIG. 2a and FIG. 2b, which will not be repeated here.

步骤S504，判断所录入的图像是否有效，如有效，输出第一图像，执行步骤S505；如无效，返回步骤S502；In step S504, it is judged whether the entered image is valid, if it is valid, the first image is output, and step S505 is executed; if it is invalid, it returns to step S502;

在本公开的一个实施例中，当获取图像后，对图像的真实性和有效性进行校验，依据预设的规则，形成的有效性验证算法，先对输入的图像进行一次初步的过滤和筛查，符合规范(主要包括图像数据的格式、分辨率要求、猪脸是否有截断等)的数据才会按照正常的流程去流转，否则会把图像数据定义为异常数据，直接退出流程，不再占用***资源进行处理。因建档过程中会采集大量数据，因此，只采用有效数据进行建档；若在使用过程中出现无效数据，则应返回数据采集部分，重新采集猪只数据以对其进行识别。In an embodiment of the present disclosure, after the image is acquired, the authenticity and validity of the image are verified. According to a preset rule, a validity verification algorithm is formed, and the input image is subjected to a preliminary filtering and summing. Screening, data that meets the specifications (mainly including the format of image data, resolution requirements, whether the pig face is truncated, etc.) will be transferred according to the normal process, otherwise the image data will be defined as abnormal data and exit the process directly. Then occupy system resources for processing. Because a large amount of data is collected during the file creation process, only valid data is used for file creation; if invalid data appears during use, you should return to the data collection section and collect pig data again to identify it.

步骤S505，确定并框画第一图像中的猪脸范围；Step S505: Determine and frame the pig face range in the first image;

在本公开的一个实施例中，完成有效性判断后，对这第一图像进行人工智能算法检测，本过程主要是对输入的第一图像进行猪脸检测，即检测出第一图像中是否存在猪脸，若存在则将猪脸框出，为后期进行识别提供支持，具体内容可参照图3，这里不再赘述。In one embodiment of the present disclosure, after the validity judgment is completed, an artificial intelligence algorithm is performed on the first image. This process is mainly to perform a pig face detection on the first input image, that is, to detect whether the first image exists Pig face, if it exists, it will be framed to provide support for later recognition. For details, please refer to Figure 3, which will not be repeated here.

步骤S506，输出框画出猪脸的第二图像；Step S506, the output frame draws a second image of a pig face;

步骤S507，对第二图像中猪脸的各部位进行标注；Step S507: annotate each part of the pig face in the second image;

在本公开的一个实施例中，将猪脸检测算法模型中输出的第二图像输入到猪脸识别算法模型中，由猪脸识别算法对检测出的猪脸进行识别，具体的，对检测出的猪脸进行打点处理，按照一定的顺序在猪脸的左鼻尖、右鼻尖、左眼睛、右眼睛、左耳根部以及右耳根部进行打点，并识别出猪脸的上述部位，识别出上述部位后，输出带有猪脸识别标记点的第三图片，同时输出一个猪脸的位置坐标向量，根据不同猪脸特征点标记位置的向量值的不同来对其进行识别，由于猪个体之间长相的差别，因此，每头猪均会输出一个唯一向量。In an embodiment of the present disclosure, the second image output from the pig face detection algorithm model is input into the pig face recognition algorithm model, and the detected pig face is recognized by the pig face recognition algorithm. Specifically, the detected The pig's face is treated with dots, and the left nose, right nose, left eye, right eye, left ear root, and right ear root of the pig face are punctured in a certain order, and the above parts of the pig face are identified, and the above parts are identified. After that, a third picture with pig face recognition mark points is output, and a pig face position coordinate vector is output at the same time, which is identified based on the difference in vector values of different pig face feature point mark positions. , So each pig will output a unique vector.

步骤S508，输出标注有猪脸各部位特征的第三图像和对应的特征向量；Step S508, output a third image labeled with features of each part of the pig face and a corresponding feature vector;

步骤S509，判断第三图像和对应的特征向量是否用于建立猪只档案，如是，执行步骤S510；如不是，执行步骤S512；Step S509, determine whether the third image and the corresponding feature vector are used to establish a pig file. If yes, go to step S510; if not, go to step S512;

步骤S510，将第三图像和对应的特征向量输入数据库；Step S510: input a third image and a corresponding feature vector into a database;

步骤S511，对第三图像和对应的特征向量所对应的猪只进行编号；Step S511: number the pigs corresponding to the third image and the corresponding feature vector;

步骤S512，将第三图像和对应的特征向量与数据库中的猪只数据进行比对；Step S512, comparing the third image and the corresponding feature vector with the pig data in the database;

步骤S513，输出对应猪只的编号；Step S513, output the number corresponding to the pig;

在本公开的一个实施例中，当猪脸识别算法输出第三图片和对应的特征向量后，判断此数据是否为建档数据，若为建档数据，则应将第三图片和对应的特征向量输入数据库中，在数据库中根据此数据建立猪只档案，并为猪只按照一定的编号顺序编制唯一编号，为后续使用过程中提供支持；若判断为不是建档数据，则应将采集数据与数据库中的猪只档案数据进行比对，寻找到此猪只数据，并将其编号输出，同时调出其相应的管理数据，以便供使用者进行处理，对工作提供帮助。In an embodiment of the present disclosure, after the pig face recognition algorithm outputs the third picture and the corresponding feature vector, it is determined whether the data is profile data. If it is profile data, the third picture and the corresponding feature should be The vector is entered into the database, and a pig file is created based on this data in the database, and a unique number is prepared for the pigs in a certain numbering sequence to provide support for subsequent use; if it is judged that it is not filed data, the collected data should be collected Compare with the pig file data in the database, find the pig data, and output its number, meanwhile call up its corresponding management data for users to process and provide work assistance.

步骤S514，结束当前流程。Step S514: End the current process.

以下介绍本公开的装置实施例，可以用于执行本公开上述的动物识别方法。The following describes device embodiments of the present disclosure, which can be used to implement the animal identification method described in the present disclosure.

图6示意性示出了根据本公开的一个实施例的动物识别装置的框图。FIG. 6 schematically illustrates a block diagram of an animal recognition device according to an embodiment of the present disclosure.

参照图6所示，根据本公开的一个实施例的动物识别装置600，包括：校验模块601、识别模块602、特征识别模块603、输出模块604；其中，Referring to FIG. 6, an animal identification device 600 according to an embodiment of the present disclosure includes: a verification module 601, an identification module 602, a feature identification module 603, and an output module 604;

校验模块601，用于对所获取的图像进行有效性校验，将无效图像滤除，获得第一图像；A verification module 601, configured to perform validity verification on the acquired image, filter out invalid images, and obtain a first image;

识别模块602，用于对第一图像中的动物面部进行识别，将所识别出的动物面部进行标识，获得标识出动物面部的第二图像；A recognition module 602, configured to recognize an animal face in a first image, identify the identified animal face, and obtain a second image identifying the animal face;

特征识别模块603，用于对第二图像进行动物面部特征识别，将所识别出的动物面部特征进行标注，形成第三图像，并基于所识别出的动物面部特征生成对应的特征向量；A feature recognition module 603, configured to perform animal facial feature recognition on the second image, mark the identified animal facial features, form a third image, and generate a corresponding feature vector based on the identified animal facial features;

输出模块604，用于基于第三图像和第三图像对应的特征向量实现动物识别。An output module 604 is configured to implement animal recognition based on the third image and the feature vector corresponding to the third image.

由于本公开的示例实施例的动物识别装置的各个功能模块与上述动物识别方法的示例实施例的步骤对应，因此对于本公开装置实施例中未披露的细节，请参照本公开上述的动物识别方法的实施例。Since each functional module of the animal identification device of the exemplary embodiment of the present disclosure corresponds to the steps of the above-mentioned example embodiment of the animal identification method, for details not disclosed in the embodiment of the device of the present disclosure, please refer to the above-mentioned animal identification method of the present disclosure. The examples.

下面参考图7，其示出了适于用来实现本公开实施例的电子设备的计算机***700的结构示意图。图7示出的电子设备的计算机***700仅是一个示例，不应对本公开实施例的功能和使用范围带来任何限制。Reference is now made to FIG. 7, which illustrates a schematic structural diagram of a computer system 700 suitable for implementing an electronic device according to an embodiment of the present disclosure. The computer system 700 of the electronic device shown in FIG. 7 is only an example, and should not impose any limitation on the functions and scope of use of the embodiments of the present disclosure.

如图7所示，计算机***700包括中央处理单元(CPU)701，其可以根据存储在只读存储器(ROM)702中的程序或者从存储部分708加载到随机访问存储器(RAM)703中的程序而执行各种适当的动作和处理。在RAM 703中，还存储有***操作所需的各种程序和数据。CPU 701、ROM 702以及RAM 703通过总线704彼此相连。输入/输出(I/O)接口705也连接至总线704。As shown in FIG. 7, the computer system 700 includes a central processing unit (CPU) 701, which can be loaded into a random access memory (RAM) 703 from a program stored in a read-only memory (ROM) 702 or from a storage section 708. Instead, perform various appropriate actions and processes. In the RAM 703, various programs and data required for system operation are also stored. The CPU 701, ROM 702, and RAM 703 are connected to each other through a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

以下部件连接至I/O接口705：包括键盘、鼠标等的输入部分1206；包括诸如阴极射线管(CRT)、液晶显示器(LCD)等以及扬声器等的输出部分707；包括硬盘等的存储部分708；以及包括诸如LAN卡、调制解调器等的网络接口卡的通信部分709。通信部分709经由诸如因特网的网络执行通信处理。驱动器710也根据需要连接至I/O接口705。可拆卸介质711，诸如磁盘、光盘、磁光盘、半导体存储器等等，根据需要安装在驱动器710上，以便于从其上读出的计算机程序根据需要被安装入存储部分708。The following components are connected to the I / O interface 705: an input portion 1206 including a keyboard, a mouse, and the like; an output portion 707 including a cathode ray tube (CRT), a liquid crystal display (LCD), and a speaker; a storage portion 708 including a hard disk and the like And a communication section 709 including a network interface card such as a LAN card, a modem, and the like. The communication section 709 performs communication processing via a network such as the Internet. The driver 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 710 as needed, so that a computer program read out therefrom is installed into the storage section 708 as needed.

特别地，根据本公开的实施例，上文参考流程图描述的过程可以被实现为计算机软件程序。例如，本公开的实施例包括一种计算机程序产品，其包括承载在计算机可读介质上的计算机程序，该计算机程序包含用于执行流程图所示的方法的程序代码。在这样的实施例中，该计算机程序可以通过通信部分709从网络上被下载和安装，和/或从可拆卸介质711被安装。在该计算机程序被中央处理单元(CPU)701执行时，执行本申请的***中限定的上述功能。In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart may be implemented as a computer software program. For example, embodiments of the present disclosure include a computer program product including a computer program carried on a computer-readable medium, the computer program containing program code for performing a method shown in a flowchart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 709, and / or installed from a removable medium 711. When this computer program is executed by a central processing unit (CPU) 701, the above-mentioned functions defined in the system of the present application are executed.

需要说明的是，本公开所示的计算机可读介质可以是计算机可读信号介质或者计算机可读存储介质或者是上述两者的任意组合。计算机可读存储介质例如可以是——但不限于——电、磁、光、电磁、红外线、或半导体的***、装置或器件，或者任意以上的组合。计算机可读存储介质的更具体的例子可以包括但不限于：具有一个或多个导线的电连接、便携式计算机磁盘、硬盘、随机访问存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、光纤、便携式紧凑磁盘只读存储器(CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。在本公开中，计算机可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行***、装置或者器件使用或者与其结合使用。而在本公开中，计算机可读的信号介质可以包括在基带中或者作为载波一部分传播的数据信号，其中承载了计算机可读的程序代码。这种传播的数据信号可以采用多种形式，包括但不限于电磁信号、光信号或上述的任意合适的组合。计算机可读的信号介质还可以是计算机可读存储介质以外的任何计算机可读介质，该计算机可读介质可以发送、传播或者传输用于由指令执行***、装置或者器件使用或者与其结合使用的程序。计算机可读介质上包含的程序代码可以用任何适当的介质传输，包括但不限于：无线、电线、光缆、RF等等，或者上述的任意合适的组合。It should be noted that the computer-readable medium shown in the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the foregoing. The computer-readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable Programming read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the foregoing. In this disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in combination with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium may include a data signal that is included in baseband or propagated as part of a carrier wave, and which carries computer-readable program code. Such a propagated data signal may take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. The computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, and the computer-readable medium may send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device . Program code embodied on a computer-readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

附图中的流程图和框图，图示了按照本公开各种实施例的***、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上，流程图或框图中的每个方框可以代表一个模块、程序段、或代码的一部分，上述模块、程序段、或代码的一部分包含一个或多个用于实现规定的逻辑功能的可执行指令。也应当注意，在有些作为替换的实现中，方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如，两个接连地表示的方框实际上可以基本并行地执行，它们有时也可以按相反的顺序执行，这依所涉及的功能而定。也要注意的是，框图或流程图中的每个方框、以及框图或流程图中的方框的组合，可以用执行规定的功能或操作的专用的基于硬件的***来实现，或者可以用专用硬件与计算机指令的组合来实现。The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, which contains one or more of the logic functions used to implement the specified logic. Executable instructions. It should also be noted that in some alternative implementations, the functions noted in the blocks may also occur in a different order than those marked in the drawings. For example, two successively represented boxes may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram or flowchart, and combinations of blocks in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with A combination of dedicated hardware and computer instructions.

描述于本公开实施例中所涉及到的单元可以通过软件的方式实现，也可以通过硬件的方式来实现，所描述的单元也可以设置在处理器中。其中，这些单元的名称在某种情况下并不构成对该单元本身的限定。The units described in the embodiments of the present disclosure may be implemented by software or hardware. The described units may also be provided in a processor. The names of these units do not, in some cases, define the unit itself.

作为另一方面，本申请还提供了一种计算机可读介质，该计算机可读介质可以是上述实施例中描述的电子设备中所包含的；也可以是单独存在，而未装配入该电子设备中。上述计算机可读介质承载有一个或者多个程序，当上述一个或者多个程序被一个该电子设备执行时，使得该电子设备实现如上述实施例中的屏幕控制实现及展示方法。As another aspect, the present application also provides a computer-readable medium, which may be included in the electronic device described in the foregoing embodiments; or may exist alone without being assembled into the electronic device in. The computer-readable medium carries one or more programs, and when the one or more programs are executed by one of the electronic devices, the electronic device is enabled to implement a screen control implementation and display method as in the above embodiments.

例如，上述的电子设备可以实现如图1中所示的：步骤S110，获取人工检票界面图像；步骤S120，对第一区域图像进行解析，获得模糊值，基于模糊值判断第一区域图像是否为车票图片；步骤S130，对第二区域图像进行解析，确定出第二区域图像的主导颜色，基于主导颜色确定出票证状态；步骤S140，当确定第一区域图像为车票图片并且票证状态为相符时，对述第三区域图像进行解析，获得证件图像；步骤S150，将证件图像与摄像头所采集的人脸图像进行匹配，获得匹配结果，基于匹配结果进行检票。For example, the above electronic device may implement as shown in FIG. 1: step S110, obtaining an image of a manual ticket checking interface; step S120, analyzing the first area image to obtain a blur value, and determining whether the first area image is based on the blur value Ticket picture; step S130, analyzing the second area image to determine the dominant color of the second area image, and determining the ticket status based on the dominant color; step S140, when it is determined that the first area image is a ticket image and the ticket status is consistent , Analyzing the third area image to obtain a credential image; step S150, matching the credential image with a face image collected by a camera to obtain a matching result, and checking the ticket based on the matching result.

又如，上述的电子设备可以实现如图4所示的各个步骤。For another example, the above electronic device can implement each step shown in FIG. 4.

又如，上述的电子设备可以实现如图7所示的各个步骤。As another example, the above electronic device can implement each step shown in FIG. 7.

应当注意，尽管在上文详细描述中提及了用于动作执行的设备的若干模块或者单元，但是这种划分并非强制性的。实际上，根据本公开的实施方式，上文描述的两个或更多模块或者单元的特征和功能可以在一个模块或者单元中具体化。反之，上文描述的一个模块或者单元的特征和功能可以进一步划分为由多个模块或者单元来具体化。It should be noted that although several modules or units of the device for action execution are mentioned in the detailed description above, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above may be embodied in one module or unit. Conversely, the features and functions of a module or unit described above can be further divided into multiple modules or units to be embodied.

通过以上的实施方式的描述，本领域的技术人员易于理解，这里描述的示例实施方式可以通过软件实现，也可以通过软件结合必要的硬件的方式来实现。因此，根据本公开实施方式的技术方案可以以软件产品的形式体现出来，该软件产品可以存储在一个非易失性存储介质(可以是CD-ROM，U盘，移动硬盘等)中或网络上，包括若干指令以使得一台计算设备(可以是个人计算机、服务器、触控终端、或者网络设备等)执行根据本公开实施方式的方法。Through the description of the above embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solution according to the embodiment of the present disclosure may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (which may be a CD-ROM, a U disk, a mobile hard disk, etc.) or on a network It includes several instructions to enable a computing device (which may be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the embodiment of the present disclosure.

本领域技术人员在考虑说明书及实践这里公开的发明后，将容易想到本公开的其它实施方案。本申请旨在涵盖本公开的任何变型、用途或者适应性变化，这些变型、用途或者适应性变化遵循本公开的一般性原理并包括本公开未公开的本技术领域中的公知常识或惯用技术手段。说明书和实施例仅被视为示例性的，本公开的真正范围和精神由下面的权利要求指出。Those skilled in the art will readily contemplate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that conform to the general principles of this disclosure and include the common general knowledge or conventional technical means in the technical field not disclosed by this disclosure. . It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

应当理解的是，本公开并不局限于上面已经描述并在附图中示出的精确结构，并且可以在不脱离其范围进行各种修改和改变。本公开的范围仅由所附的权利要求来限制。It should be understood that the present disclosure is not limited to the precise structure that has been described above and illustrated in the accompanying drawings, and various modifications and changes may be made without departing from the scope thereof. The scope of the disclosure is limited only by the following claims.

Claims

一种动物识别方法，其特征在于，包括：An animal identification method, comprising:

对所获取的图像进行有效性校验，将无效图像滤除，获得第一图像；Perform validity check on the acquired image, filter out invalid images, and obtain a first image;

对所述第一图像中的动物面部进行识别，将所识别出的动物面部进行标识，获得标识出动物面部的第二图像；Identifying the animal face in the first image, identifying the identified animal face, and obtaining a second image identifying the animal face;

对所述第二图像进行动物面部特征识别，将所识别出的动物面部特征进行标注，形成第三图像，并基于所识别出的动物面部特征生成对应的特征向量；Performing animal facial feature recognition on the second image, labeling the identified animal facial features, forming a third image, and generating a corresponding feature vector based on the identified animal facial features;

基于所述第三图像和所述第三图像对应的特征向量实现动物识别。Animal recognition is achieved based on the third image and a feature vector corresponding to the third image.
根据权利要求1所述的动物识别方法，其特征在于，所获取的图像至少包括所述图像的数据格式信息、分辨率信息和动物面部；所述对所获取的图像进行有效性校验，将无效图像滤除，获得第一图像包括：The animal identification method according to claim 1, wherein the acquired image includes at least data format information, resolution information, and animal face of the image; and performing validity check on the acquired image, Filtering out invalid images and obtaining the first image includes:

将所述图像输入预设的有效性校验模型，获得所述图像的有效性校验结果；Inputting the image into a preset validity check model to obtain a validity check result of the image;

基于所述有效性校验结果，将不符合预设数据格式和预设分辨率的图像滤除，获得过滤后的图像；Filtering out images that do not conform to a preset data format and a preset resolution based on the validity check result to obtain a filtered image;

将所述过滤后的图像中的动物面部不完整的图像滤除，获得第一图像。The incomplete image of the animal face in the filtered image is filtered to obtain a first image.
根据权利要求2所述的动物识别方法，其特征在于，所述方法还包括：The animal identification method according to claim 2, wherein the method further comprises:

当判断所述图像不符合预设数据格式和预设分辨率，和\或判断所述图像中动物面部不完整时，生成并发送重新采集图像的控制信息。When it is determined that the image does not conform to a preset data format and a preset resolution, and / or it is determined that an animal face in the image is incomplete, control information for reacquiring the image is generated and sent.
根据权利要求1所述的动物识别方法，其特征在于，对所述第一图像中的动物面部进行识别，将所识别出的动物面部进行标识，获得标识出动物面部的第二图像，包括：The animal identification method according to claim 1, wherein identifying the animal face in the first image, identifying the identified animal face, and obtaining a second image identifying the animal face, comprising:

将所述第一图像输入至预设的动物面部识别模型，获得与所述第一图像对应的动物面部识别结果；Inputting the first image to a preset animal face recognition model to obtain an animal face recognition result corresponding to the first image;

基于所述动物面部识别结果，确定所述第一图像中动物面部范围；Determining an animal face range in the first image based on the animal face recognition result;

将所述动物面部范围框画在方框内，生成框画出动物面部范围的第二图像。Draw the frame of the animal face within the frame, and generate a second image that draws the frame of the animal face.
根据权利要求1所述的动物识别方法，其特征在于，对所述第二图像进行动物面部特征识别，将所识别出的动物面部特征进行标注，形成第三图像，包括：The animal identification method according to claim 1, wherein performing animal facial feature recognition on the second image, and labeling the identified animal facial features to form a third image, comprising:

识别出所述第二图像动物面部的各部位；Identifying parts of the animal face in the second image;

按照预设的顺序对所识别出的各个部位进行标注，获得带有动物面部的各个部位标注的第三图像。Annotate each identified part in a preset order to obtain a third image with annotated parts of the animal's face.
根据权利要求5所述的动物识别方法，其特征在于，所述各部位至少包括：眼睛部位、鼻子部位、耳朵部位；所述按照预设的顺序对所识别出的各个部位进行标注，获得带有动物面部的各个部位标注的第三图像，包括：The animal identification method according to claim 5, wherein each part includes at least: an eye part, a nose part, and an ear part; and each of the identified parts is marked in a preset order to obtain a band A third image labeled with various parts of the animal's face, including:

在所述第二图像中所识别出眼睛部位的左眼和右眼标注出位置点；The left eye and the right eye of the eye part identified in the second image are marked with position points;

在所述第二图像中所识别出鼻子部位的左鼻尖和右鼻尖标注出位置点；The left nose tip and the right nose tip of the nose part identified in the second image are marked with position points;

在所述第二图像中所识别出耳朵部位的左耳根部和右耳根部标注出位置点；Mark the location points of the left ear root and the right ear root of the ear part identified in the second image;

获得有标注出所述眼睛部位左眼和右眼的位置点、所述鼻子部位左鼻尖和右鼻尖的位置点、所述耳朵部位左耳根部和右耳根的位置点的第三图像。A third image is obtained with the positions of the left and right eyes of the eye part, the positions of the left and right nose tip of the nose part, and the positions of the left and right ear base of the ear part.
根据权利要求6所述的动物识别方法，其特征在于，所述基于所识别出的动物面部特征生成对应的特征向量，包括：The animal identification method according to claim 6, wherein the generating a corresponding feature vector based on the identified facial features of the animal comprises:

根据所述第三图像中动物面部的眼睛部位左眼和右眼的位置点、所述鼻子部位左鼻尖和右鼻尖的位置点、所述耳朵部位左耳根部和右耳根的位置点与所述框画出动物面部范围方框的相对位置，生成所述动物面部的特征向量。According to the position points of the left and right eyes of the eye part of the animal face in the third image, the position points of the left and right nose tips of the nose part, and the position points of the left and right ear roots of the ear part and the The frame draws the relative position of the animal face range box, and generates a feature vector of the animal face.
一种动物识别装置，其特征在于，包括：校验模块、识别模块、特征识别模块、输出模块；其中，An animal identification device, comprising: a verification module, an identification module, a feature identification module, and an output module; wherein,

校验模块，用于对所获取的图像进行有效性校验，将无效图像滤除，获得第一图像；A verification module, configured to verify the validity of the acquired image, filter out invalid images, and obtain a first image;

识别模块，用于对所述第一图像中的动物面部进行识别，将所识别出的动物面部进行标识，获得标识出动物面部的第二图像；A recognition module, configured to recognize an animal face in the first image, identify the identified animal face, and obtain a second image identifying the animal face;

特征识别模块，用于对所述第二图像进行动物面部特征识别，将所识别出的动物面部特征进行标注，形成第三图像，并基于所识别出的动物面部特征生成对应的特征向量；A feature recognition module, configured to perform animal facial feature recognition on the second image, mark the identified animal facial features, form a third image, and generate a corresponding feature vector based on the identified animal facial features;

输出模块，用于基于所述第三图像和所述第三图像对应的特征向量实现动物识别。An output module is configured to implement animal recognition based on the third image and a feature vector corresponding to the third image.
一种计算机可读介质，其上存储有计算机程序，其特征在于，所述程序被处理器执行时实现权利要求1至7中任一项所述的动物识别方法。A computer-readable medium having stored thereon a computer program, characterized in that when the program is executed by a processor, the animal identification method according to any one of claims 1 to 7 is implemented.
一种电子设备，其特征在于，包括：An electronic device, comprising:

一个或多个处理器；One or more processors;

存储装置，用于存储一个或多个程序，当所述一个或多个程序被所述一个或多个处理器执行时，使得所述一个或多个处理器实现权利要求1至7中任一所述的动物识别方法。A storage device, configured to store one or more programs, and when the one or more programs are executed by the one or more processors, cause the one or more processors to implement any one of claims 1 to 7 The animal identification method.