Biological diversity image data induction method, equipment and medium

Abstract

The invention provides a biological diversity image data induction method and device and a medium, and the method comprises the following steps: obtaining a target parameter based on a target demand, obtaining a target image set according to the target parameter, obtaining a recognizable image set based on the target image set, inputting a sample image into a target recognition model, and obtaining a recognition image set based on the recognizable image set; the method comprises the steps of obtaining a sample image, obtaining a target weight corresponding to the sample image, determining whether the sample image is a recognizable image or not based on the target weight, summarizing a recognizable image set, and storing each type of recognizable image into a special folder. According to the method, the parameters of the infrared camera are set by utilizing multi-factor correlation analysis so as to meet the requirement of scene shooting, and the shot photos are screened, so that the utilization rate of effective photos is improved, and the accuracy of concluding the biodiversity image data is relatively high.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a method, device and medium for summarizing biodiversity image data. Background Art

[0002] As a non-destructive sampling technology, infrared camera technology has many advantages such as all-weather uninterrupted operation, strong concealment, little interference to wild animals, and easy preservation and retrieval of image data. These characteristics make infrared camera technology widely used in biodiversity monitoring. Infrared cameras play a vital role in obtaining biodiversity image data. The amount of biodiversity image data captured by infrared cameras is huge. Summarizing the acquired image data is an important task. It helps to efficiently store, manage and analyze a large amount of biological image data, and plays an important role in biodiversity research and ecological protection. Therefore, how to accurately summarize the huge amount of biodiversity data has become an urgent problem to be solved. Effective summarization of biodiversity image data can better provide new perspectives and insights for biodiversity research. Summary of the invention

[0003] In view of the above technical problems, the technical solution adopted by the present invention is: a method for summarizing biodiversity image data, comprising the following steps:

[0004] S100, based on target requirements, obtaining target parameters, wherein the target parameters are relevant parameters of the camera set so that the photos taken by the infrared camera meet the target number, and the target number is the number of photos obtained by each continuous shooting of the target area.

[0005] S200, acquiring a target image set according to target parameters, wherein the target image set includes a plurality of target images, and the target images are images acquired by photographing a target area with an infrared camera set with target parameters.

[0006] S300, based on the target image set, obtaining a recognizable image set, wherein the recognizable image set includes a plurality of recognizable images, wherein in S300, the recognizable images are obtained by the following steps:

[0007] S301, input the sample image into the target recognition model to obtain the target weight λ corresponding to the sample image, where λ meets the following conditions:

[0008] λ=∑ f e＝1 η e / f,η e is the probability that the target recognition model recognizes the e-th organism in the sample image, e=1…f, and f is the number of organisms recognized in the sample image.

[0009] S302, when λ≥λ 0 , the sample image is determined to be a recognizable image, where λ 0 is the preset probability threshold.

[0010] S400, summarize the recognizable image set and store each type of recognizable image into a dedicated folder.

[0011] The present invention protects a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned method for summarizing biodiversity image data when executing the computer program.

[0012] The present invention protects a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the computer program implements the above-mentioned method for summarizing biodiversity image data.

[0013] The present invention has at least the following beneficial effects: the present invention is a method for summarizing biodiversity image data, the method comprising the following steps: based on target requirements, obtaining target parameters, according to the target parameters, obtaining a target image set, based on the target image set, obtaining a recognizable image set, wherein a sample image is input into a target recognition model, a target weight corresponding to the sample image is obtained, based on the target weight, it is determined whether the sample image is a recognizable image, the recognizable image set is summarized, and each type of recognizable image is stored in a special folder. Based on the requirements for each category of images, the present invention uses multi-factor correlation analysis to set the parameters of the infrared camera to meet the requirements for scene shooting, screens the taken photos, improves the utilization rate of effective photos, and thereby makes the accuracy of summarizing the biodiversity image data higher. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0015] Figure 1 A flowchart of a method for summarizing biodiversity image data provided by an embodiment of the present invention;

[0016] Figures 2 to 5 A schematic diagram of a scenario to which a method for summarizing biodiversity image data provided by an embodiment of the present invention is applicable. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0018] Example

[0019] This embodiment provides a method for summarizing biodiversity image data, the method comprising the following steps: Figures 1 to 5 As shown:

[0020] S100, based on target requirements, obtaining target parameters, wherein the target parameters are relevant parameters of the camera set so that the photos taken by the infrared camera meet the target number, and the target number is the number of photos obtained by each continuous shooting of the target area.

[0021] Specifically, in S100, the target parameters are obtained through the following steps:

[0022] S101, obtain a sample image list set A = {A 1 , ..., A i , ..., A n}, A i ={A i1 , ..., A ij , ..., A im}, A ij is the jth sample image list corresponding to the ith sample area, j=1...m, m is the number of sample image lists corresponding to the ith sample area, i=1...n, n is the number of sample areas.

[0023] Specifically, the sample image list includes several sample images, wherein the sample image is any image in the continuous shooting image obtained by photographing the sample area with an infrared camera with set candidate parameters, wherein the candidate parameters are relevant parameters of the camera set so that the photos taken by the infrared camera meet the candidate quantity, and the candidate quantity is the number of photos obtained in each continuous shooting.

[0024] Furthermore, those skilled in the art know that any method in the prior art for setting infrared camera parameters according to the number of continuous shots falls within the protection scope of the present invention.

[0025] Furthermore, the sample area is an area where the shooting is performed, for example, a sample area such as a wild ecological environment area, a zoo area, etc.

[0026] S102, according to A, obtain the candidate quantity list B corresponding to the candidate parameter list = {B 1 , ..., B j , ..., B m} and the candidate priority list set F corresponding to B = {F 1 , ..., F j , ..., F m}, F j ={F j1 , ..., F jr , ..., F js}, B j is the number of candidates corresponding to the jth candidate parameter, F j For B j The corresponding candidate priority is the rth candidate priority in the candidate priority list, where r=1…s, and s is the number of candidate priorities in the candidate priority list.

[0027] Preferably, the value of s is 3.

[0028] Specifically, in S102, F is obtained by the following steps: jr :

[0029] S1021, get B j The corresponding target image quantity list set D j ={D 1 j ,……,D r j ,……,D s j}, D r j ={D r j1 ,……,D r ji ,……,D r jn}, D r ji For B j The number of target images of the rth category corresponding to the i-th sample area.

[0030] Specifically, the number of target images is the number of target images, wherein the target images are images of a certain type obtained by classifying sample images according to preset categories.

[0031] Furthermore, the preset categories are preset categories based on the biometric recognition rate in the image, wherein those skilled in the art know that the preset categories can be set according to actual needs, all of which fall within the protection scope of the present invention and will not be repeated here. For example, when s=3, the sample images are divided into three categories, namely, recognizable images, aerial images, and blurred images. Therefore, the number of target images includes the number of recognizable images in the sample images, the number of aerial images in the sample images, and the number of blurred images in the sample images, wherein the recognizable images are images that can identify a certain organism, the aerial images are images whose probability of biological recognition is lower than a certain threshold, and the blurred images are images whose probability of biological recognition is between two thresholds.

[0032] S1022, according to D j , get F jr , where F jr Meet the following conditions:

[0033] F jr =β r / ∑ s r＝1 β r ,in, |B j -D r ji |For(B j -D r ji ), and ε is a constant between 0 and 1.

[0034] Specifically, those skilled in the art know that ε can be selected according to actual needs, all of which fall within the protection scope of the present invention and will not be described in detail here.

[0035] S103, obtaining a target requirement, wherein the target requirement is a requirement for the number of images of each type when shooting with an infrared camera.

[0036] S104, based on F and target requirements, obtaining target parameters, wherein the target parameters are relevant parameters of the camera set so that the photos taken by the infrared camera meet the target number, the target number is the number of photos obtained in each continuous shooting to meet the target requirements based on the target neural network model, and the target neural network model is a neural network model obtained by inputting F and the target requirements as training data into the initial neural network model for training.

[0037] Specifically, those skilled in the art know that the initial neural network model can be selected according to actual needs, which falls within the protection scope of the present invention and will not be described in detail here.

[0038] S200, acquiring a target image set according to target parameters, wherein the target image set includes a plurality of target images, and the target images are images acquired by photographing a target area with an infrared camera set with target parameters.

[0039] Specifically, the target area is an area where a number of biodiversity images are to be captured for summarization, for example, a wild ecological environment area, a zoo area, and other target areas.

[0040] S300, based on the target image set, obtaining a recognizable image set, wherein the recognizable image set includes a plurality of recognizable images, wherein in S300, the recognizable images are obtained by the following steps:

[0041] S301, input the sample image into the target recognition model to obtain the target weight λ corresponding to the sample image, where λ meets the following conditions:

[0042] λ=∑ f e＝1 η e / f,η e is the probability that the target recognition model recognizes the e-th organism in the sample image, e=1…f, and f is the number of organisms recognized in the sample image.

[0043] S302, when λ≥λ 0 , the sample image is determined to be a recognizable image, where λ 0 is the preset probability threshold.

[0044] Specifically, 0 The value range of λ is 0.6 to 0.8, wherein those skilled in the art know that λ can be adjusted according to actual needs. 0 The selection of all falls within the protection scope of the present invention and will not be repeated here.

[0045] S400, summarize the recognizable image set and store each type of recognizable image into a dedicated folder.

[0046] Specifically, the types of images that can be identified include: people, wild animals, poultry, and pets.

[0047] Specifically, after step S400, the following steps are also included:

[0048] Perform face authentication and expression recognition on the current frame image;

[0049] S500, obtaining a set of images to be selected, wherein the set of images to be selected includes a plurality of images to be selected, and the images to be selected are recognizable images of a person selected from a set of recognizable images in chronological order.

[0050] S600, when the face authentication result and expression recognition of each candidate image in the candidate image set are authenticated according to the preset rules, confirm that the identity authentication of this face is passed, wherein the preset rules are that the number of frames in tracking is greater than a first preset threshold and the result of dividing the number of frames that pass the authentication by the number of frames in tracking is greater than a second preset threshold and the number of expressions is not less than 2.

[0051] Specifically, before performing face authentication and expression recognition on the current frame image, it also includes: performing face detection and tracking on the current frame image. If tracked, it is the same face and continues to determine whether the previous frame image has passed identity authentication. If the identity authentication is passed, information that the identity authentication has passed is output; if the identity authentication is not passed, the current frame image is performed face authentication and expression recognition; if not tracked, it is a different face and the relevant data of the previous face comprehensive authentication is cleared, and the current frame image is performed face authentication and expression recognition.

[0052] Specifically, the face identity authentication is performed by assuming that the person is a real person rather than a photo, wherein the amount of data calculated is small, less system resources are occupied, and a real person and a photo can be quickly distinguished, thus meeting the needs of real-time applications.

[0053] Specifically, those skilled in the art know that the first preset threshold and the second preset threshold can be selected according to actual needs, which all fall within the protection scope of the present invention and will not be described in detail here.

[0054] Specifically, the preset rule may also be that M consecutive frames of images pass face authentication, and the expression recognition result is the first expression, and N consecutive frames of images after the Mth frame pass face authentication, and the expression recognition result is the second expression, wherein M and N are natural numbers.

[0055] As described above, based on the demand for each category of images, multi-factor correlation analysis is used to set the parameters of the infrared camera to meet the demand for scene shooting, and the taken photos are screened, which improves the utilization rate of effective photos, thereby making the accuracy of summarizing biodiversity image data higher.

[0056] The present embodiment provides a method for summarizing biodiversity image data, the method comprising the following steps: based on target requirements, obtaining target parameters, obtaining a target image set according to the target parameters, and obtaining a recognizable image set based on the target image set, wherein a sample image is input into a target recognition model, a target weight corresponding to the sample image is obtained, and based on the target weight, it is determined whether the sample image is a recognizable image, the recognizable image set is summarized, and each type of recognizable image is stored in a special folder. Based on the requirements for each category of images, the present invention uses multi-factor correlation analysis to set the parameters of the infrared camera to meet the requirements for scene shooting, and screens the taken photos, thereby improving the utilization rate of effective photos, thereby making the summarization of biodiversity image data more accurate.

[0057] An embodiment of the present invention also provides a non-transitory computer-readable storage medium, which can be set in an electronic device to store at least one instruction or at least one program related to implementing a method in a method embodiment. The at least one instruction or the at least one program is loaded and executed by the processor to implement the method provided in the above embodiment.

[0058] An embodiment of the present invention further provides an electronic device, comprising a processor and the aforementioned non-transitory computer-readable storage medium.

[0059] Although some specific embodiments of the present invention have been described in detail by way of example, it will be appreciated by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the present invention. It will also be appreciated by those skilled in the art that various modifications may be made to the embodiments without departing from the scope and spirit of the present invention. The scope of the present invention is defined by the appended claims.

Claims

1. A method for summarizing biodiversity image data, characterized in that: The method comprises the following steps: S100, based on the target requirement, obtaining target parameters, wherein the target parameters are relevant parameters of the camera set so that the photos taken by the infrared camera meet the target number, and the target number is the number of photos obtained by each continuous shooting of the target area; S200, acquiring a target image set according to the target parameters, wherein the target image set includes a plurality of target images, and the target images are images acquired by photographing a target area with an infrared camera set with the target parameters; S300, based on the target image set, obtaining a recognizable image set, wherein the recognizable image set includes a plurality of recognizable images, wherein in S300, the recognizable images are obtained by the following steps: S301, input the sample image into the target recognition model to obtain the target weight λ corresponding to the sample image, where λ meets the following conditions: λ=∑ f e＝1 η e / f,η e is the probability that the target recognition model recognizes the e-th organism in the sample image, e=1…f, f is the number of organisms recognized in the sample image; S302, when λ≥λ 0 , the sample image is determined to be a recognizable image, where λ 0 is the preset probability threshold; S400, summarize the recognizable image set, and store each type of recognizable image into a dedicated folder.

2. The method for summarizing biodiversity image data according to claim 1, characterized in that: In S100, the target parameters are obtained through the following steps: S101, obtain a sample image list set A = {A1, ..., A i , ..., A n }, A i ={A i1 , ..., A ij , ..., A im }, A ij is the jth sample image list corresponding to the i-th sample area, j=1...m, m is the number of sample image lists corresponding to the i-th sample area, i=1...n, n is the number of sample areas; S102, according to A, obtain a candidate quantity list B corresponding to the candidate parameter list = {B1, ..., B j , ..., B m } and B corresponding to the candidate priority list set F = {F1, ..., F j , ..., F m }, F j ={F j1 , ..., F jr , ..., F js }, B j is the number of candidates corresponding to the jth candidate parameter, F j For B j The rth candidate priority in the corresponding candidate priority list, r=1…s, s is the number of candidate priorities in the candidate priority list; S103, obtaining a target requirement, wherein the target requirement is a requirement for the number of images of each type when shooting with an infrared camera; S104, based on F and target requirements, obtaining target parameters, wherein the target parameters are relevant parameters of the camera set so that the photos taken by the infrared camera meet the target number, the target number is the number of photos obtained in each continuous shooting to meet the target requirements based on the target neural network model, and the target neural network model is a neural network model obtained by inputting F and the target requirements as training data into the initial neural network model for training.

3. The method for summarizing biodiversity image data according to claim 2, characterized in that: The sample image list includes several sample images, wherein the sample image is any image in the continuous shooting images obtained by photographing the sample area with an infrared camera with set candidate parameters, wherein the candidate parameters are relevant parameters of the camera set so that the photos taken by the infrared camera meet the candidate quantity, and the candidate quantity is the number of photos obtained in each continuous shooting.

4. The method for summarizing biodiversity image data according to claim 2, characterized in that: In S102, F is obtained by the following steps: jr : S1021, get B j The corresponding target image number list set D j ={D 1 j ,……,D r j ,……,D s j }, D r j ={D r j1 ,……,D r ji ,……,D r jn }, D r ji For B j The number of target images of the rth category corresponding to the corresponding i-th sample area; S1022, according to D j , get F jr , where F jr Meet the following conditions: F jr =b r / ∑ s r＝1 b r , among them, |B j -D r ji (B j -D r ji ) The absolute value of , ε is a constant between 0 and 1.

5. The method for summarizing biodiversity image data according to claim 4, characterized in that: The target image quantity is the number of target images, wherein the target images are images of a certain type obtained by classifying sample images according to preset categories.

6. The method for summarizing biodiversity image data according to claim 5, characterized in that: The preset categories are preset categories divided based on the biometric recognition rate in the image.

7. A non-transitory computer-readable storage medium, characterized in that: The storage medium stores at least one instruction or at least one program, and the at least one instruction or the at least one program is loaded and executed by the processor to implement the method as claimed in any one of claims 1 to 6.

8. An electronic device, characterized in that: The invention comprises a processor and the non-transitory computer-readable storage medium as claimed in claim 7.

Images