CN111597911A - Method and system for rapidly extracting key frame based on image characteristics - Google Patents
Method and system for rapidly extracting key frame based on image characteristics Download PDFInfo
- Publication number
- CN111597911A CN111597911A CN202010321923.2A CN202010321923A CN111597911A CN 111597911 A CN111597911 A CN 111597911A CN 202010321923 A CN202010321923 A CN 202010321923A CN 111597911 A CN111597911 A CN 111597911A
- Authority
- CN
- China
- Prior art keywords
- image
- frame
- characteristic
- change value
- image feature
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method and a system for rapidly extracting key frames based on image characteristics, wherein the method comprises the following steps: s1: acquiring image characteristics of a first frame image, and setting and storing an image identifier as 1; s2: acquiring image characteristics of the next frame of image; s3: acquiring an image characteristic change value; s4, judging whether the image characteristic change value exceeds a preset image characteristic change threshold value, if so, configuring the current frame image characteristic as a characteristic template, setting the saved image identifier as 1, and executing the step S2; if not, go to step S5; s5: judging whether the saved image identifier is 1, if so, saving the current frame image, setting the saved image identifier to be 0, and executing the step S2, wherein the current frame image is a key frame image; if not, go to step S2. The aim of quickly extracting the key frame is achieved by comparing the image characteristics of the front frame and the back frame, setting a preset threshold value and storing an image identifier, meanwhile, the action frame is filtered, the workload is reduced, and a large amount of time overhead is saved.
Description
Technical Field
The invention relates to the technical field of human-computer interaction, in particular to a method and a system for quickly extracting key frames based on image features.
Background
The current key frame extraction algorithm, such as the K-means algorithm, needs to give the number of categories in advance, needs to continuously perform sample classification adjustment, and continuously calculates new adjusted clustering centers, so that the time overhead of the algorithm is very large when the data volume is very large.
Disclosure of Invention
In view of the foregoing prior art, a first object of the present invention is to provide a method for quickly extracting a key frame based on image features, in which only image features of a current frame are compared with a feature template, so as to reduce the amount of calculation, save a large amount of time overhead, and implement filtering of a motion frame and extraction of a key frame.
The invention is realized by the following technical scheme:
a method for rapidly extracting key frames based on image features comprises the following steps:
s1: acquiring image characteristics of a first frame image, configuring the image characteristics of the first frame image as a characteristic template, and setting and storing an image identifier as 1;
s2: acquiring image characteristics of a next frame image, wherein the next frame image is a current frame image, and the next frame image characteristics are configured as the current frame image characteristics;
s3: obtaining an image characteristic change value obtained by comparing the current frame image characteristic with the characteristic template;
s4: judging whether the image characteristic change value exceeds a preset image characteristic change threshold value, if so, updating the current frame image characteristic as a characteristic template, converting and storing the image identifier as 1, and executing a step S2; if not, go to step S5;
s5: judging whether the saved image identifier is 1, if so, saving the current frame image, wherein the current frame image is a key frame image, converting the saved image identifier to be 0, and executing a step S2; if not, go to step S2.
Optionally, the image feature comprises an image feature shape and an image feature point.
Further, the image feature shape includes a convex hull shape, a concave hull shape, a contour shape of a circumscribed rectangle, and a contour shape of a circumscribed circle.
Further, the image feature points comprise convex wrap points, concave wrap points, a central point of a circumscribed rectangle and the circle center of a circumscribed circle.
Further, the image feature change threshold includes an area threshold of the image feature shape and a coordinate threshold of the image feature point; the image feature change value includes an area change value of the image feature shape and a coordinate change value of the image feature point.
A second object of the present invention is to provide a system for rapidly extracting a key frame based on image features, comprising:
the preset module is used for presetting an image characteristic change threshold;
the setting module is used for setting the stored image identifier as 1 when the first frame image is displayed, converting the stored image identifier and updating the characteristic template;
the acquisition module is used for acquiring image characteristics and image characteristic change values;
the judging module is used for judging whether the image characteristic change value exceeds the image characteristic change threshold value and judging whether the saved image identifier is 1;
and the storage module is used for storing the key frames.
Optionally, the image feature comprises an image feature shape and an image feature point.
Further, the image feature shape includes a convex hull shape, a concave hull shape, a contour shape of a circumscribed rectangle, and a contour shape of a circumscribed circle.
Further, the image feature points comprise convex wrap points, concave wrap points, a central point of a circumscribed rectangle and the circle center of a circumscribed circle.
Further, the image feature change threshold includes an area threshold of the image feature shape and a coordinate threshold of the image feature point; the image feature change value includes an area change value of the image feature shape and a coordinate change value of the image feature point.
Compared with the prior art, the invention has the following advantages and beneficial effects:
the invention provides a method for rapidly extracting key frames based on image characteristics, which judges the similarity between the image characteristics of a current frame and a characteristic template by comparing the image characteristics of the current frame with the characteristic template and setting an image characteristic change threshold, reduces the workload and saves a large amount of time overhead; by converting and storing the image identifiers, images with small similarity are stored to achieve the purpose of quickly extracting key frames, meanwhile, action frames of continuous similar images are filtered, and redundancy of the key frames is reduced.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:
fig. 1 is a schematic flowchart of a method for rapidly extracting a key frame based on image features according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a key frame extraction process in a method for quickly extracting key frames based on image features according to an embodiment of the present invention
Fig. 3 is a schematic structural diagram of a system for rapidly extracting a key frame based on image features according to a second embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to examples and accompanying drawings, and the exemplary embodiments and descriptions thereof are only used for explaining the present invention and are not meant to limit the present invention.
The existing key frame extraction method usually adopts a K-means algorithm, the K-means algorithm is an indirect clustering method based on similarity measurement among samples, the algorithm takes K as a parameter and divides n objects into K clusters, so that the clusters have higher similarity and the similarity among the clusters is lower. The algorithm first randomly selects k objects and for each of the remaining objects, assigns it to the most similar cluster based on the distance between the object and the cluster center. Then, a new center for each cluster is calculated and the process is repeated. That is, the K-means algorithm needs to give the number of categories in advance, needs to continuously perform sample classification adjustment, and continuously calculates new adjusted cluster centers, so that when the data size is very large, the calculation amount is large, and the time overhead of the algorithm is very large.
Therefore, the present invention provides an improved method for extracting key frames based on image features, which aims at the defects of the existing method for extracting key frames based on the K-means algorithm, and it is known that, under the condition that a shot does not have sudden change, the image information feature value of each frame of a continuous video is in a slightly changed state, the image information feature value of any frame in a video sequence is always close to the image information feature values of the frames before and after the shot, that is, for a non-sudden video sequence, the feature value of each frame of the video sequence is gradually changed, and when the gradual change is accumulated to a certain extent, the feature values of some frames in the video sequence are compared, and the feature values may have been greatly changed. The invention only compares the image characteristics of the front and back frames of images by utilizing the characteristic, thereby completing the real-time judgment of all the frames, realizing the filtration of continuous similar motion frames, and storing the image with larger change after gradual change accumulation as the key frame.
Example one
The embodiment provides a method for quickly extracting a key frame based on image features, as shown in fig. 1, including the following steps:
s1: acquiring image characteristics of a first frame image, configuring the image characteristics of the first frame image as a characteristic template, and setting and storing an image identifier as 1;
in this embodiment, the first frame image is used as a start frame, the image features of the first frame image are extracted, the image features of the first frame image are used as a feature template, and the saved image identifier is set to 1.
It should be noted that the setting of the start frame is not limited to the first frame image, and the start frame may be set according to the user-defined key frame to extract the video segment.
S2: acquiring the image characteristics of the next frame of image, wherein the next frame of image is the current frame of image, and the image characteristics of the next frame of image are configured as the current frame of image characteristics;
s3: obtaining an image characteristic change value obtained by comparing the current frame image characteristic with a characteristic template;
s4: judging whether the image characteristic change value exceeds a preset image characteristic change threshold value, if so, updating the current frame image characteristic as a characteristic template, converting and saving the image identifier to be 1, and executing a step S2; if not, go to step S5;
s5: judging whether the stored image identifier is 1, if so, storing the current frame image, wherein the current frame image is a key frame image, converting the stored image identifier to be 0, and executing the step S2; if not, go to step S2.
The invention provides a method for rapidly extracting key frames based on image characteristics, which comprises the steps of firstly comparing the image characteristics of a current frame with the image characteristics (namely a characteristic template) of a first frame, and judging whether an image characteristic change value obtained after comparison exceeds a preset image characteristic change threshold value or not; if the current frame image characteristic is not similar to the characteristic template, the current frame image characteristic is not stored, the current frame image characteristic is updated to be the characteristic template, the stored image identifier is converted into 1, and then the step of obtaining the image characteristic of the next frame image is repeated.
If the image feature of the current frame is not over, the image feature of the current frame is similar to the feature template, meanwhile, whether the storage identifier is 1 or not needs to be judged, if the image feature change value is not over the preset image feature change threshold and the identifier of the current storage image is 1, the current frame image is stored, the current frame image is a key frame, the storage image identifier is converted into 0, and then the step of obtaining the image feature of the next frame image is repeated; if the image characteristic change value does not exceed the preset image characteristic change threshold value and the current saved image identifier is 0, the current frame image is one of continuous similar images, namely the image is not saved, and the characteristic template is not updated at the same time, and then the step of obtaining the image characteristics of the next frame image is repeated until the real-time judgment of all the frames and the extraction of the key frames are completed. The invention judges the similarity between the image characteristics of the current frame and the characteristic template by comparing the image characteristics of the current frame with the characteristic template and setting an image characteristic change threshold, because the characteristic value of each frame of image in the video sequence is gradually changed, the characteristic template is continuously updated along with the accumulation of the image characteristic value, thus realizing the real-time judgment of the image characteristics and the characteristic template of all frames, reducing the calculated amount, saving a large amount of time expenditure, saving the image with smaller similarity by converting and storing the image identifier, achieving the purpose of quickly extracting the key frame, and simultaneously realizing the filtering of the action frame of continuous similar images.
Wherein the image features include image feature shapes and image feature points. The image characteristic shapes comprise convex hull shapes, concave hull shapes, outline shapes of circumscribed rectangles and outline shapes of circumscribed circles, and the image characteristic points comprise convex hull points, concave hull points, central points of the circumscribed rectangles and circle centers of the circumscribed circles; in addition, the convex hull shape and the like are used as the contour features, the convex hull points and the like are used as the feature points, and the variation between the front frame image and the rear frame image of the video is convenient to distinguish.
It should be noted that the image feature shape is not limited to the convex hull shape, the concave hull shape, the outline shape of the circumscribed rectangle, and the outline shape of the circumscribed circle, and other shapes can be used as the shape for identifying the image feature; the image feature points are not limited to the convex wrap points, the concave wrap points, the center points of the circumscribed rectangles and the centers of the circumscribed circles, and other points can be used for identifying the image features.
The image characteristic change threshold comprises an area threshold of an image characteristic shape and a coordinate threshold of an image characteristic point; the image feature change value includes an area change value of an image feature shape and a coordinate change value of an image feature point. Specifically, when the similarity between the previous and subsequent frames of the video is determined, that is, when step S4 is executed, the area change value and the coordinate change value of the image feature point, and the area threshold and the coordinate threshold of the corresponding image feature shape are determined, and an image with a smaller similarity is determined by setting the threshold, so as to achieve the purpose of quickly extracting the key frame.
In this embodiment, since the image feature value of each frame in the video is in a state of a micro-variation, when the micro-variation is accumulated to a certain degree, the feature value will change greatly, as shown in fig. 2, the implementation process of applying the method for rapidly extracting the key frame based on the image feature of the present invention to the key frame extraction is as follows:
1) acquiring a video for monitoring a human body in real time from a camera, presetting an area change threshold of an image characteristic shape as 100 pixels and a coordinate change threshold of an image characteristic point as 10 pixels, extracting an image characteristic of a first frame as a characteristic template, and setting and storing an image identifier as 1;
2) when acquiring the image characteristics of the second frame image, comparing the second frame with the first frame image: the area change value is 90, the coordinate change value is 3, the area change value and the coordinate change value do not exceed a preset threshold, the image identifier is stored to be 1, the second frame image is stored, and meanwhile the image identifier is converted to be 0;
3) when the image characteristics of the third frame image are acquired, comparing the third frame image with the first frame image: the area change value is 93, the coordinate change value is 5, the area change value and the coordinate change value do not exceed a preset threshold, the image identifier is saved to be 0, and the third frame image is not saved;
4) when the image characteristics of the fourth frame image are acquired, comparing the fourth frame image with the first frame image: the area change value is 98, the coordinate change value is 9, the area change value and the coordinate change value do not exceed the preset threshold, the image identifier is saved to be 0, and the fourth frame image is not saved;
5) when the image characteristics of the image of the fifth frame are acquired, comparing the image of the fifth frame with the image of the first frame: the area change value is 101, the coordinate change value is 12, the area change value and the coordinate change value exceed a preset threshold value, the fourth frame image is not stored, the fifth frame image characteristic is updated to be a characteristic template, and meanwhile, the stored image identifier is converted into 1;
6) when the image characteristics of the sixth frame image are acquired, comparing the sixth frame image with the fifth frame image: the area change value is 90, the coordinate change value is 2, the area change value and the coordinate change value do not exceed a preset threshold, the image identifier is stored to be 1, the sixth frame of image is stored, and meanwhile the image identifier is converted to be 0;
7) when the image characteristics of the seventh frame image are acquired, comparing the seventh frame image with the fifth frame image: the area change value is 92, the coordinate change value is 3, the area change value and the coordinate change value do not exceed a preset threshold, the image identifier is saved to be 0, and the seventh frame image is not saved;
8) when the image characteristics of the image of the eighth frame are acquired, comparing the image of the eighth frame with the image of the fifth frame: the area change value is 94, the coordinate change value is 5, the area change value and the coordinate change value do not exceed the preset threshold, the image identifier is saved to be 0, and the seventh frame image is not saved;
and repeating the steps until the real-time judgment of all the frames and the extraction of the key frames are completed. The similarity between frames is judged by comparing the image characteristics of the current frame with the characteristic template, setting a threshold value and storing an image identifier, images with smaller similarity (such as the second frame and the sixth frame of images in fig. 2) are stored to achieve the purpose of quickly extracting key frames, continuous similar images (such as the third frame, the fourth frame and the fifth frame of images in fig. 2) are filtered out, the redundancy of the key frames is reduced, and a large amount of time overhead is saved.
It should be noted that, in practical application, since the first frame image is used as an initial frame for feature extraction, it is impossible to determine whether the first frame of the video is a valid frame, and only the first frame of the video is valid by default, and an identification algorithm needs to be input for identification, and all the subsequent images can be extracted as a key frame in a motion process through threshold adjustment.
It should be noted that, in practical applications, the change value of each frame image in the video is not limited to the above 1) -8), and corresponding key frame extraction and motion frame screening may be obtained according to practical situations.
Example two
The embodiment provides a system for quickly extracting a key frame based on image features, as shown in fig. 2, including:
the preset module is used for presetting an image characteristic change threshold;
the setting module is used for setting and saving the image identifier;
the acquisition module is used for acquiring image characteristics and image characteristic change values;
the judging module is used for judging whether the image characteristic change value exceeds an image characteristic change threshold value and judging whether the saved image identifier is 1;
and the storage module is used for storing the key frames.
According to the system for rapidly extracting the key frame based on the image characteristics, provided by the invention, the similarity between the image characteristics of the current frame and the characteristic template is judged by comparing the image characteristics of the current frame with the characteristic template and setting the image characteristic change threshold, so that the workload is reduced, and a large amount of time overhead is saved; by converting and storing the image identifiers, images with small similarity are stored to achieve the purpose of quickly extracting key frames, and meanwhile, moving frames of continuous similar images are filtered out, so that the redundancy of the key frames is reduced. The specific working process of the system for quickly extracting a key frame based on image features is as described in the first embodiment of the method for quickly extracting a key frame based on image features, and is not described herein again.
Wherein the image features include image feature shapes and image feature points. The image characteristic shapes comprise convex hull shapes, concave hull shapes, outline shapes of circumscribed rectangles and outline shapes of circumscribed circles, and the image characteristic points comprise convex hull points, concave hull points, central points of the circumscribed rectangles and circle centers of the circumscribed circles; in addition, the convex hull shape and the like are used as the contour features, the convex hull points and the like are used as the feature points, and the variation between the front frame image and the rear frame image of the video is convenient to distinguish.
It should be noted that the image feature shape is not limited to the convex hull shape, the concave hull shape, the outline shape of the circumscribed rectangle, and the outline shape of the circumscribed circle, and other shapes can be used as the shape for identifying the image feature; the image feature points are not limited to the convex wrap points, the concave wrap points, the center points of the circumscribed rectangles and the centers of the circumscribed circles, and other points can be used for identifying the image features.
The image characteristic change threshold comprises an area threshold of an image characteristic shape and a coordinate threshold of an image characteristic point; the image feature change value includes an area change value of an image feature shape and a coordinate change value of an image feature point. Specifically, when the similarity between the previous and subsequent frames of the video is determined, that is, when step S4 is executed, the area change value and the coordinate change value of the image feature point, and the area threshold and the coordinate threshold of the corresponding image feature shape are determined, and an image with a smaller similarity is determined by setting the threshold, so as to achieve the purpose of quickly extracting the key frame.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (10)
1. A method for rapidly extracting key frames based on image features is characterized by comprising the following steps:
s1: acquiring image characteristics of a first frame image, configuring the image characteristics of the first frame image as a characteristic template, and setting and storing an image identifier as 1;
s2: acquiring image characteristics of a next frame image, wherein the next frame image is a current frame image, and the next frame image characteristics are configured as the current frame image characteristics;
s3: obtaining an image characteristic change value obtained by comparing the current frame image characteristic with the characteristic template;
s4: judging whether the image characteristic change value exceeds a preset image characteristic change threshold value, if so, updating the current frame image characteristic as a characteristic template, converting and storing the image identifier as 1, and executing a step S2; if not, go to step S5;
s5: judging whether the saved image identifier is 1, if so, saving the current frame image, wherein the current frame image is a key frame image, converting the saved image identifier to be 0, and executing a step S2; if not, go to step S2.
2. The method for rapidly extracting key frames based on image features as claimed in claim 1, wherein the image features comprise image feature shapes and image feature points.
3. The method for rapidly extracting key frames based on image features as claimed in claim 2, wherein the image feature shapes comprise convex hull shapes, concave hull shapes, outline shapes of circumscribed rectangles and outline shapes of circumscribed circles.
4. The method for rapidly extracting key frames based on image features as claimed in claim 2, wherein the image feature points comprise convex points, concave points, circumscribed rectangle center points and the centers of circumscribed circles.
5. The method for rapidly extracting key frames based on image features according to claim 2, wherein the image feature change threshold comprises an area threshold of the image feature shape and a coordinate threshold of the image feature point; the image feature change value includes an area change value of the image feature shape and a coordinate change value of the image feature point.
6. A system for rapidly extracting key frames based on image features is characterized by comprising:
the preset module is used for presetting an image characteristic change threshold;
the setting module is used for setting the stored image identifier as 1 when the first frame image is displayed, converting the stored image identifier and updating the characteristic template;
the acquisition module is used for acquiring image characteristics and image characteristic change values;
the judging module is used for judging whether the image characteristic change value exceeds the image characteristic change threshold value and judging whether the saved image identifier is 1;
and the storage module is used for storing the key frames.
7. The system for rapidly extracting key frames based on image features as claimed in claim 6, wherein the image features comprise image feature shapes and image feature points.
8. The system for rapidly extracting key frames based on image features as claimed in claim 7, wherein the image feature shapes comprise convex hull shapes, concave hull shapes, outline shapes of circumscribed rectangles and outline shapes of circumscribed circles.
9. The system for rapidly extracting key frames based on image features as claimed in claim 7, wherein the image feature points comprise convex points, concave points, center points of circumscribed rectangles and centers of circumscribed circles.
10. The system for rapidly extracting key frames based on image features as claimed in claim 7, wherein the image feature change threshold comprises an area threshold of the image feature shape and a coordinate threshold of the image feature point; the image feature change value includes an area change value of the image feature shape and a coordinate change value of the image feature point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010321923.2A CN111597911B (en) | 2020-04-22 | 2020-04-22 | Method and system for rapidly extracting key frames based on image features |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010321923.2A CN111597911B (en) | 2020-04-22 | 2020-04-22 | Method and system for rapidly extracting key frames based on image features |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111597911A true CN111597911A (en) | 2020-08-28 |
| CN111597911B CN111597911B (en) | 2023-08-29 |
Family
ID=72181634
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010321923.2A Active CN111597911B (en) | 2020-04-22 | 2020-04-22 | Method and system for rapidly extracting key frames based on image features |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111597911B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112613436A (en) * | 2020-12-28 | 2021-04-06 | 中国联合网络通信集团有限公司 | Examination cheating detection method and device |
| CN112906818A (en) * | 2021-03-17 | 2021-06-04 | 东南数字经济发展研究院 | Method for reducing redundancy of video data set during artificial intelligence training |
| CN113158914A (en) * | 2021-04-25 | 2021-07-23 | 胡勇 | Intelligent evaluation method for dance action posture, rhythm and expression |
| CN114786052A (en) * | 2022-04-29 | 2022-07-22 | 同方知网数字出版技术股份有限公司 | Academic live video fast browsing method based on key frame extraction |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120027295A1 (en) * | 2009-04-14 | 2012-02-02 | Koninklijke Philips Electronics N.V. | Key frames extraction for video content analysis |
| US20160247024A1 (en) * | 2015-02-23 | 2016-08-25 | Kodak Alaris Inc. | Method for output creation based on video content characteristics |
| WO2017000465A1 (en) * | 2015-07-01 | 2017-01-05 | 中国矿业大学 | Method for real-time selection of key frames when mining wireless distributed video coding |
| US20170076078A1 (en) * | 2014-05-12 | 2017-03-16 | Ho Kim | User authentication method, device for executing same, and recording medium for storing same |
| CN107480580A (en) * | 2017-03-31 | 2017-12-15 | 触景无限科技（北京）有限公司 | Image-recognizing method and pattern recognition device |
| KR101870700B1 (en) * | 2017-03-07 | 2018-06-25 | 광운대학교 산학협력단 | A fast key frame extraction method for 3D reconstruction from a handheld video |
| CN110427517A (en) * | 2019-07-18 | 2019-11-08 | 华戎信息产业有限公司 | A kind of figure based on scene lexicographic tree searches video method, device and computer readable storage medium |
| CN110929605A (en) * | 2019-11-11 | 2020-03-27 | 中国建设银行股份有限公司 | Video key frame storage method, device, equipment and storage medium |
| CN112446363A (en) * | 2021-01-29 | 2021-03-05 | 广州市玄武无线科技股份有限公司 | Image splicing and de-duplication method and device based on video frame extraction |
| US20210223046A1 (en) * | 2018-12-13 | 2021-07-22 | Goertek Inc. | Method and device for extracting key frames in simultaneous localization and mapping and smart device |
-
2020
- 2020-04-22 CN CN202010321923.2A patent/CN111597911B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120027295A1 (en) * | 2009-04-14 | 2012-02-02 | Koninklijke Philips Electronics N.V. | Key frames extraction for video content analysis |
| US20170076078A1 (en) * | 2014-05-12 | 2017-03-16 | Ho Kim | User authentication method, device for executing same, and recording medium for storing same |
| US20160247024A1 (en) * | 2015-02-23 | 2016-08-25 | Kodak Alaris Inc. | Method for output creation based on video content characteristics |
| WO2017000465A1 (en) * | 2015-07-01 | 2017-01-05 | 中国矿业大学 | Method for real-time selection of key frames when mining wireless distributed video coding |
| KR101870700B1 (en) * | 2017-03-07 | 2018-06-25 | 광운대학교 산학협력단 | A fast key frame extraction method for 3D reconstruction from a handheld video |
| CN107480580A (en) * | 2017-03-31 | 2017-12-15 | 触景无限科技（北京）有限公司 | Image-recognizing method and pattern recognition device |
| US20210223046A1 (en) * | 2018-12-13 | 2021-07-22 | Goertek Inc. | Method and device for extracting key frames in simultaneous localization and mapping and smart device |
| CN110427517A (en) * | 2019-07-18 | 2019-11-08 | 华戎信息产业有限公司 | A kind of figure based on scene lexicographic tree searches video method, device and computer readable storage medium |
| CN110929605A (en) * | 2019-11-11 | 2020-03-27 | 中国建设银行股份有限公司 | Video key frame storage method, device, equipment and storage medium |
| CN112446363A (en) * | 2021-01-29 | 2021-03-05 | 广州市玄武无线科技股份有限公司 | Image splicing and de-duplication method and device based on video frame extraction |
Non-Patent Citations (4)
| Title |
|---|
| 宁煜西等: "民航航班跟踪视频关键帧提取方法研究", 《空军预警学院学报》 * |
| 宁煜西等: "民航航班跟踪视频关键帧提取方法研究", 《空军预警学院学报》, no. 03, 15 June 2018 (2018-06-15) * |
| 郭忠伟等: "C~3I系统中战场视频镜头检测与关键帧提取", 《火力与指挥控制》 * |
| 郭忠伟等: "C~3I系统中战场视频镜头检测与关键帧提取", 《火力与指挥控制》, 15 June 2008 (2008-06-15) * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112613436A (en) * | 2020-12-28 | 2021-04-06 | 中国联合网络通信集团有限公司 | Examination cheating detection method and device |
| CN112613436B (en) * | 2020-12-28 | 2023-08-04 | 中国联合网络通信集团有限公司 | Examination cheating detection method and device |
| CN112906818A (en) * | 2021-03-17 | 2021-06-04 | 东南数字经济发展研究院 | Method for reducing redundancy of video data set during artificial intelligence training |
| CN113158914A (en) * | 2021-04-25 | 2021-07-23 | 胡勇 | Intelligent evaluation method for dance action posture, rhythm and expression |
| CN114786052A (en) * | 2022-04-29 | 2022-07-22 | 同方知网数字出版技术股份有限公司 | Academic live video fast browsing method based on key frame extraction |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111597911B (en) | 2023-08-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111597911B (en) | Method and system for rapidly extracting key frames based on image features | |
| CN108710865B (en) | Driver abnormal behavior detection method based on neural network | |
| CN104778396B (en) | A kind of eyeprint identification unlocking method and system based on environmental screening frame | |
| CN107092884B (en) | Rapid coarse-fine cascade pedestrian detection method | |
| CN110706235B (en) | Far infrared pedestrian detection method based on two-stage cascade segmentation | |
| CN109472262A (en) | Licence plate recognition method, device, computer equipment and storage medium | |
| CN114926436A (en) | Defect detection method for periodic pattern fabric | |
| CN108369644B (en) | Method for quantitatively detecting human face raised line, intelligent terminal and storage medium | |
| WO2017092272A1 (en) | Face identification method and device | |
| Lili et al. | The algorithm of iris image preprocessing | |
| CN114494318B (en) | Cornea contour extraction method based on cornea dynamic deformation video of Ojin algorithm | |
| CN110222647B (en) | Face in-vivo detection method based on convolutional neural network | |
| CN107633251A (en) | A kind of vehicle identification system based on image enhaucament | |
| CN117173416B (en) | Railway freight train number image definition processing method based on image processing | |
| CN116824526B (en) | Digital intelligent road monitoring system based on image processing | |
| CN112699901A (en) | Plant species identification system based on Internet of things | |
| CN110971826A (en) | Video front-end monitoring device and method | |
| KR101669447B1 (en) | System and the method for recognizing drowsiness | |
| Saranya et al. | An approach towards ear feature extraction for human identification | |
| CN112418085B (en) | Facial expression recognition method under partial shielding working condition | |
| CN114299586A (en) | Intelligent deep learning system based on convolutional neural network | |
| CN111950409B (en) | Intelligent identification method and system for road marking line | |
| KR100893086B1 (en) | Method for detecting face robust to illumination change | |
| CN111488889A (en) | Intelligent image processor for image edge extraction | |
| CN110781810A (en) | Face emotion recognition method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |