KR-20260066869-A - Method and apparatus for detecting abnormal situations using real-time video data

KR20260066869AKR 20260066869 AKR20260066869 AKR 20260066869AKR-20260066869-A

Abstract

The present invention relates to a method and apparatus for detecting abnormal situations using real-time video data. A method for detecting abnormal situations using real-time video data according to one aspect may include: receiving a frame of the real-time video data and preprocessing it; inputting the preprocessed frame into a first feature extraction model according to a preset method to output a first feature; adding the output first feature to a feature queue; inputting at least one feature among a plurality of features included in the feature queue into a second feature extraction model to output a second feature; and detecting an abnormal situation based on an abnormality score calculated based on the second feature.

Inventors

강민성
임영철

Assignees

재단법인대구경북과학기술원

Dates

Publication Date: 20260512
Application Date: 20241105

Claims (11)

In a method for detecting abnormal situations using real-time video data, A step of receiving and preprocessing frames of the above real-time video data; A step of inputting the above-mentioned preprocessed frame into a first feature extraction model according to a preset method to output a first feature; A step of adding the above-mentioned first feature to a feature queue; A step of inputting features included in the above feature queue into a second feature extraction model to output a second feature; and A step of detecting an abnormal situation based on an abnormal score calculated based on the second feature above; A method including
In Article 1, The step of outputting the above-mentioned first feature is, When the number of frames reaches N, the method includes the step of inputting the N frames into a first feature extraction model to output a first feature; A method in which N is a natural number greater than or equal to 2.
In Article 2, A method in which the above N is set by user input.
In Article 1, The step of adding to the above feature queue is, A step of adding the first feature output above to the feature queue, and deleting the feature that was added to the feature queue first among a plurality of features included in the feature queue; A method including
In Article 1, A method in which the maximum number of features included in the above feature queue is set by user input.
In Article 1, The step of detecting the above abnormal situation is, A step of inputting the output second feature into a feature correction model to output a corrected second feature; and A step of calculating an abnormality score based on the above-mentioned corrected second feature and detecting an abnormal situation based on the above-mentioned abnormality score; A method including
In Article 1, The above abnormal score is, A method calculated based on the difference between a predefined normal pattern and the second feature.
In Article 1, The above method is, A step of receiving training information of the first feature extraction model and the second feature extraction model; A method that further includes.
In Article 1, The above preprocessing step is, A method performed in parallel with at least one of the steps of outputting the first feature, adding to the feature queue, outputting the second feature, and detecting the abnormal situation.
A computer-readable recording medium having a program for executing the method of claim 1 on a computer.
At least one memory; and It includes at least one processor, The above processor is, Receive and preprocess the frames of the above real-time video data, The above-mentioned preprocessed frame is input into a first feature extraction model according to a preset method to output a first feature, and Add the above-mentioned first feature to the feature queue, and At least one feature among the plurality of features included in the above feature queue is input into a second feature extraction model to output a second feature, and Detecting an abnormal situation based on an abnormal score calculated based on the second feature above, A device for detecting abnormal situations using real-time video data.

Description

Method and apparatus for detecting abnormal situations using real-time video data The present disclosure relates to a method and apparatus for detecting abnormal situations using real-time video data. To detect anomalies from video, there are technologies that detect objects in video data and use spatial information of those objects to detect anomalies. However, detecting anomalies using only spatial information of objects may result in lower accuracy in anomaly judgment compared to using temporal information, such as context between frames, in conjunction. To overcome these limitations, using both spatial and temporal information simultaneously for anomaly detection can increase data throughput and computational burden. In particular, when attempting to detect anomalies from real-time video data, data throughput and computational burden can be further exacerbated. With the recent advancement of deep learning technology, it has become possible to effectively process spatiotemporal information using artificial intelligence models, and attempts are being made to detect abnormal situations from real-time video data using artificial intelligence models. The aforementioned background technology is technical information that the inventor possessed for the derivation of the present invention or acquired during the process of deriving the present invention, and it cannot be considered as prior art disclosed to the general public prior to the filing of the present invention. FIG. 1 is a diagram illustrating an example of a process for performing anomaly detection from real-time video data according to one embodiment. FIG. 2 is a flowchart illustrating a method for detecting abnormal situations using real-time video data according to one embodiment. FIG. 3 is a conceptual diagram illustrating a method for extracting a first feature according to one embodiment. FIG. 4 is a conceptual diagram illustrating a method for extracting a second feature according to one embodiment. FIG. 5 is a conceptual diagram illustrating the process of performing a method for detecting abnormal situations through parallel processing threads according to one embodiment. FIG. 6 is an illustrative diagram for explaining the results of abnormal situation detection according to one embodiment. FIG. 7 is a block diagram of a user device according to one embodiment. The advantages and features of the present invention, and the methods for achieving them, will become clear by referring to the embodiments described in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments presented below, but can be implemented in various different forms and should be understood to include all modifications, equivalents, and substitutions that fall within the spirit and scope of the present invention. The embodiments presented below are provided to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention. In describing the present invention, detailed descriptions of related known technologies are omitted if it is determined that such detailed descriptions may obscure the essence of the present invention. The terms used in the embodiments have been selected to be as close as possible to currently widely used general terms; however, these may vary depending on the intent of those skilled in the art, case law, the emergence of new technologies, etc. Additionally, in specific cases, terms have been selected at the applicant's discretion, and in such cases, their meanings will be described in detail in the relevant description section. Therefore, terms used in the specification must be defined not merely by their names, but based on their meanings and the content throughout the specification. When a part of a specification is described as 'comprising' a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components. Furthermore, terms such as 'unit' and 'module' used as suffixes in the specification refer to a unit that handles at least one function or operation, and this may be implemented in hardware or software, or as a combination of hardware and software. Additionally, terms including ordinal numbers, such as 'first' or 'second', may be used in the specification to describe various components, but said components shall not be limited by said terms. Such terms may be used for the purpose of distinguishing one component from another. Some embodiments of the present disclosure may be represented by functional block configurations and various processing steps. Some or all of these functional blocks may be implemented by various numbers of hardware and/or software configurations that execute specific functions. For example, the functional blocks of the present disclosure may be implemented by one or more microprocessors or by circuit configurations