CN-115690674-B - Data compatible processing method based on smart campus and Internet of things cloud platform

CN115690674BCN 115690674 BCN115690674 BCN 115690674BCN-115690674-B

Abstract

The invention provides a data compatible processing method based on an intelligent campus and an Internet of things cloud platform, and relates to the technical field of intelligent campus and data processing. The method comprises the steps of carrying out data compatible conversion operation on a plurality of campus Internet of things monitoring videos acquired and formed through a plurality of campus Internet of things terminal devices to form a plurality of corresponding target campus Internet of things monitoring videos, carrying out video frame splicing processing on target campus Internet of things monitoring video frames included in the plurality of target campus Internet of things monitoring videos based on corresponding video frame timestamp information to form corresponding spliced campus Internet of things monitoring videos, and carrying out video identification processing on the spliced campus Internet of things monitoring videos by utilizing a video identification neural network formed through network optimization to output target video identification results corresponding to the spliced campus Internet of things monitoring videos. Based on the above, the problem of low efficiency of data compatible processing in the prior art can be improved.

Inventors

CHEN ZHIXIONG
Xu Zhuqiong

Assignees

广州市威士丹利智能科技有限公司

Dates

Publication Date: 20260505
Application Date: 20221012
Priority Date: 20220928

Claims (9)

1. The data compatible processing method based on the intelligent campus is characterized by being applied to an Internet of things cloud platform, and comprises the following steps: Performing data compatible conversion operation on a plurality of campus Internet of things monitoring videos acquired and formed by a plurality of campus Internet of things terminal devices to form a plurality of corresponding target campus Internet of things monitoring videos, wherein the plurality of target campus Internet of things monitoring videos have the same video format; Extracting each frame of target campus internet of things monitoring video frame with the video frame time stamp information from the plurality of target campus internet of things monitoring videos aiming at each video frame time stamp information based on the video frame time stamp information corresponding to each frame of target campus internet of things monitoring video frame included in each target campus internet of things monitoring video so as to form a video frame set to be spliced corresponding to the video frame time stamp information; Aiming at each piece of video frame time stamp information, performing video frame splicing processing on each target campus Internet of things monitoring video frame included in a video frame set to be spliced corresponding to the video frame time stamp information according to the position relation among corresponding intelligent campus subregions to form a corresponding spliced campus Internet of things monitoring video frame, combining the spliced campus Internet of things monitoring video frames according to the corresponding video frame time stamp information to form a corresponding spliced campus Internet of things monitoring video, wherein the spliced campus Internet of things monitoring video comprises multi-frame spliced campus Internet of things monitoring video frames, and each spliced campus Internet of things monitoring video frame is formed by splicing multi-frame target campus Internet of things monitoring video frames with the same video frame time stamp information; And carrying out video identification processing on the spliced campus Internet of things monitoring video by utilizing a video identification neural network formed by network optimization so as to output a target video identification result corresponding to the spliced campus Internet of things monitoring video, wherein the target video identification result is used for reflecting the campus security state of the intelligent campus area corresponding to the spliced campus Internet of things monitoring video.
2. The method for processing data compatibility based on intelligent campus as claimed in claim 1, wherein the step of performing data compatibility conversion operation on the plurality of campus internet of things monitoring videos collected and formed by the plurality of campus internet of things terminal devices to form a plurality of corresponding target campus internet of things monitoring videos comprises: Transmitting a synchronous data acquisition instruction to each campus Internet of things terminal device in a plurality of campus Internet of things terminal devices in communication connection, so that each campus Internet of things terminal device synchronously acquires data of a corresponding intelligent campus subarea according to the synchronous data acquisition instruction to form a plurality of synchronous campus Internet of things monitoring videos, wherein any two-frame campus Internet of things monitoring video frames with the same video frame time sequence have the same video frame time stamp information; and performing data compatible conversion operation on the plurality of campus Internet of things monitoring videos to form a plurality of corresponding target campus Internet of things monitoring videos.
3. The smart campus-based data compatible processing method as claimed in claim 1 or 2, wherein the step of performing video identification processing on the spliced campus internet of things monitoring video by using a video identification neural network formed by performing network optimization to output a target video identification result corresponding to the spliced campus internet of things monitoring video comprises: Performing video feature mining operation on the spliced campus internet of things monitoring video by utilizing a video feature mining sub-network included in a video identification neural network formed by performing network optimization, and outputting a video feature data mining result corresponding to the spliced campus internet of things monitoring video; Utilizing a video object feature mining sub-network included in the video identification neural network to perform video object feature mining operation on the spliced campus Internet of things monitoring video and video object information included in a preset video object information set, and outputting a video object feature data mining result corresponding to the spliced campus Internet of things monitoring video; utilizing a video segment feature mining sub-network included in the video identification neural network to perform video segment feature mining operation on the spliced campus Internet of things monitoring video and the matched video segments included in the preset matched video segment set, and outputting a video segment feature data mining result corresponding to the spliced campus Internet of things monitoring video; aggregating the video feature data mining result, the video object feature data mining result and the video segment feature data mining result to form an aggregate data mining result corresponding to the spliced campus internet of things monitoring video; and loading the aggregate data mining result to a video recognition sub-network included in the video recognition neural network, so as to obtain a target video recognition result corresponding to the spliced campus Internet of things monitoring video by utilizing the video recognition sub-network.
4. The smart campus-based data compatible processing method as claimed in claim 3, wherein the step of performing video feature mining operation on the spliced campus internet of things surveillance video by using a video feature mining sub-network included in a video recognition neural network formed by performing network optimization, and outputting a video feature data mining result corresponding to the spliced campus internet of things surveillance video comprises: Performing video segmentation operation on the spliced campus internet of things monitoring video to form spliced campus internet of things monitoring video segments of the spliced campus internet of things monitoring video, and performing video segment mapping operation on the spliced campus internet of things monitoring video segments to output video segment mapping results corresponding to the spliced campus internet of things monitoring video segments; analyzing video clip distribution information of the spliced campus internet of things monitoring video clips in the spliced campus internet of things monitoring video, and performing distribution information mapping operation on the video clip distribution information to output a video clip distribution information mapping result corresponding to the video clip distribution information; Analyzing and outputting a video segment identification information mapping result corresponding to the spliced campus Internet of things monitoring video segment, and then performing result aggregation operation on the video segment mapping result, the video segment distribution information mapping result and the video segment identification information mapping result to form an aggregation information mapping result corresponding to the spliced campus Internet of things monitoring video segment; Loading the aggregation information mapping result into a video feature mining sub-network included in a video identification neural network formed by network optimization, performing video feature mining operation on the aggregation information mapping result by utilizing the video feature mining sub-network, outputting an initial video feature data mining result corresponding to the spliced campus Internet of things monitoring video segment, and analyzing and outputting a video feature data mining result corresponding to the spliced campus Internet of things monitoring video according to the initial video feature data mining result corresponding to the spliced campus Internet of things monitoring video segment.
5. The smart campus-based data compatible processing method as claimed in claim 3, wherein the step of using the video object feature mining sub-network included in the video identification neural network to perform video object feature mining operation on the spliced campus internet of things surveillance video and video object information included in a pre-configured video object information set, and outputting a video object feature data mining result corresponding to the spliced campus internet of things surveillance video includes: performing object comparison operation on the spliced campus Internet of things monitoring video and video object information included in a preset video object information set to output an object comparison result corresponding to the spliced campus Internet of things monitoring video; When the object comparison result reflects that the spliced campus internet of things monitoring video has the video object information included in the video object information set, marking the video object information of the spliced campus internet of things monitoring video to form corresponding video object information to be processed; Loading the video object information to be processed into a video object feature mining sub-network included in the video identification neural network, so as to perform video object feature mining operation on the video object information to be processed by utilizing the video object feature mining sub-network, and outputting video object information feature distribution corresponding to the video object information to be processed; And analyzing and outputting a video object characteristic data mining result corresponding to the spliced campus internet of things monitoring video according to the video object information characteristic distribution.
6. The smart campus-based data compatible processing method as claimed in claim 3, wherein the step of using the video segment feature mining sub-network included in the video identification neural network to perform video segment feature mining operation on the spliced campus internet of things surveillance video and the matched video segments included in the pre-configured matched video segment set, and outputting a video segment feature data mining result corresponding to the spliced campus internet of things surveillance video includes: Performing video segment comparison operation on the spliced campus Internet of things monitoring video and the matched video segments included in the preset matched video segment set to output video segment comparison results corresponding to the spliced campus Internet of things monitoring video; When the video segment comparison result reflects that the matched video segments matched with the spliced campus Internet of things monitoring video exist in the matched video segment set, marking the matched video segments matched with the spliced campus Internet of things monitoring video in the matched video segment set so as to form corresponding matched video segments to be processed; Loading the to-be-processed matching video segments into a video segment feature mining sub-network included in the video identification neural network, so as to perform video segment feature mining operation on the to-be-processed matching video segments by utilizing the video segment feature mining sub-network, and outputting matching video segment feature distribution corresponding to the to-be-processed matching video segments; and analyzing and outputting video segment characteristic data mining results corresponding to the spliced campus Internet of things monitoring video according to the matched video segment characteristic distribution.
7. The smart campus-based data compatible processing method as claimed in claim 3, wherein the step of performing video recognition processing on the spliced campus internet of things surveillance video by using a video recognition neural network formed by performing network optimization to output a target video recognition result corresponding to the spliced campus internet of things surveillance video further comprises: The extracted sample spliced campus internet of things monitoring video and sample video annotation results corresponding to the sample spliced campus internet of things monitoring video are used for reflecting the real campus security state corresponding to the sample spliced campus internet of things monitoring video; performing video feature mining operation on the monitoring video of the sample spliced campus Internet of things by utilizing a to-be-optimized video feature mining sub-network included in the to-be-optimized video identification neural network, and outputting sample video feature data mining results corresponding to the monitoring video of the sample spliced campus Internet of things; Performing video object feature mining operation on video object information included in the monitoring video of the sample spliced campus Internet of things and a preset video object information set by using a to-be-optimized video object feature mining sub-network included in the to-be-optimized video identification neural network, and outputting sample video object feature data mining results corresponding to the monitoring video of the sample spliced campus Internet of things; performing video segment feature mining operation on the spliced campus internet of things monitoring video and the matched video segments included in the preset matched video segment set by utilizing a to-be-optimized video segment feature mining sub-network included in the to-be-optimized video identification neural network, and outputting an example video segment feature data mining result corresponding to the example spliced campus internet of things monitoring video; and performing network optimization operation on the to-be-optimized video recognition neural network according to the example video feature data mining result, the example video object feature data mining result, the example video fragment feature data mining result, the example video labeling result and the to-be-optimized video recognition sub-network included in the to-be-optimized video recognition neural network so as to form a video recognition neural network corresponding to the to-be-optimized video recognition neural network.
8. The smart campus-based data-compatible processing method as claimed in claim 7, wherein the step of performing a network optimization operation on the video recognition neural network to be optimized according to the example video feature data mining result, the example video object feature data mining result, the example video clip feature data mining result, the example video annotation result, and the video recognition sub-network to be optimized included in the video recognition neural network to form the video recognition neural network corresponding to the video recognition neural network to be optimized includes: Aggregating the sample video feature data mining result, the sample video object feature data mining result and the sample video fragment feature data mining result to form a sample aggregate data mining result of the sample spliced campus internet of things monitoring video, and loading the sample aggregate data mining result into a to-be-optimized video recognition sub-network included in the to-be-optimized video recognition neural network to obtain a sample video recognition result corresponding to the sample spliced campus internet of things monitoring video by utilizing the to-be-optimized video recognition sub-network recognition; according to the example video identification result and the example video annotation result, analyzing and outputting a network optimization cost value corresponding to the video identification neural network to be optimized; Under the condition that the network optimization cost value corresponding to the video identification neural network to be optimized is greater than or equal to a pre-configured network optimization cost reference value, performing network optimization operation on the video identification neural network to be optimized according to the network optimization cost value; Marking the video identification neural network to be optimized after the network optimization operation as a candidate video identification neural network, performing network optimization operation on the candidate video identification neural network, and marking the current candidate video identification neural network as a video identification neural network under the condition that the network optimization cost value corresponding to the candidate video identification neural network after the network optimization operation is smaller than the network optimization cost reference value.
9. The cloud platform of the internet of things, which is characterized by comprising a processor and a memory, wherein the memory is used for storing a computer program, and the processor is used for executing the computer program to realize the data compatible processing method based on the intelligent campus as claimed in any one of claims 1-8.

Description

Data compatible processing method based on smart campus and Internet of things cloud platform Technical Field The invention relates to the technical field of intelligent campus and data processing, in particular to a data compatible processing method based on intelligent campus and an Internet of things cloud platform. Background The intelligent campus aims at promoting the fusion of information technology and education teaching and improving the effect of learning and teaching, and aims at providing an intelligent learning environment which is comprehensively perceived, intelligent, dataized, networked and collaborative integrated in teaching, scientific research, management and life service and can perform insight and prediction on education teaching and education management by taking new technologies such as the Internet of things, cloud computing, big data analysis and the like as core technologies. Smart campus = 1 data center + smart campus infrastructure + eight types of smart campus applications + smart resources. The eight intelligent campus application systems are respectively a student growth intelligent application system, a teacher professional development intelligent application system, a scientific research intelligent application system, an education management intelligent application system, a security monitoring intelligent application system, a logistics service intelligent application system, a social service intelligent application system and a comprehensive evaluation intelligent application system. The internet of things refers to a network which is used for connecting any article with the internet through information sensing equipment according to a contracted protocol and carrying out information exchange and communication so as to realize intelligent identification, positioning, tracking, monitoring and management. In popular terms, the Internet of things is the Internet with connected things, and comprises two layers of meanings, wherein the first Internet of things is the extension and expansion of the Internet, the core and the foundation of the Internet are still the Internet, and the second Internet of things comprises people and articles, so that the Internet of things realizes the exchange and communication of information among people, articles and articles. In the prior art, data collected by different terminal devices of the internet of things are generally processed respectively, so that the problem of low efficiency of data compatibility processing exists. Disclosure of Invention Accordingly, the invention aims to provide a data compatible processing method based on an intelligent campus and an internet of things cloud platform, so as to solve the problem of low efficiency of data compatible processing in the prior art. In order to achieve the above purpose, the embodiment of the present invention adopts the following technical scheme: A data compatible processing method based on an intelligent campus is applied to an Internet of things cloud platform, and the data compatible processing method based on the intelligent campus comprises the following steps: Performing data compatible conversion operation on a plurality of campus Internet of things monitoring videos acquired and formed by a plurality of campus Internet of things terminal devices to form a plurality of corresponding target campus Internet of things monitoring videos, wherein the plurality of target campus Internet of things monitoring videos have the same video format; Video frame splicing processing is carried out on the target campus internet of things monitoring video frames included in the target campus internet of things monitoring videos based on video frame time stamp information corresponding to each frame of target campus internet of things monitoring video frame included in the target campus internet of things monitoring videos so as to form corresponding spliced campus internet of things monitoring videos, the spliced campus internet of things monitoring videos include multi-frame spliced campus internet of things monitoring video frames, and each frame of spliced campus internet of things monitoring video frame is formed by splicing multi-frame target campus internet of things monitoring video frames with the same video frame time stamp information; And carrying out video identification processing on the spliced campus Internet of things monitoring video by utilizing a video identification neural network formed by network optimization so as to output a target video identification result corresponding to the spliced campus Internet of things monitoring video, wherein the target video identification result is used for reflecting the campus security state of the intelligent campus area corresponding to the spliced campus Internet of things monitoring video. In some preferred embodiments, in the data compatible processing method based on smart campus, the step of performing a data compatible conversi