Method and system for identifying and pasting objects in video

Abstract

The invention discloses a method and a system for identifying and pasting objects in a video. The method comprises the following steps of: S1, guiding a user to upload a video containing a product photographed in the system; S2, performing intelligent analysis on the uploaded video to obtain whether the product is photographed in the video or not; S3, performing whole network information retrieval and output according to a return value in the S2; S4, performing corresponding processing on the video, and correspondingly implanting materials into the video to form a plurality of new videos implanted with the materials; S5, forming generated sharing information into content convenient to share quickly; and S6, sharing the content to social media, social software or other content media by the user. The system comprises a video uploading sub-module, a video analysis sub-module, an object identification sub-module, a video processing sub-module and a video sharing sub-module. The method and the system are high in recognition accuracy rate for the types of products appearing in the video, are high in richness of the form of implanted materials, are highly targeted, and are beneficial to propaganda.

Description

technical field [0001] The invention relates to the technical field of video processing, in particular to an object recognition in video, a patch method and a system thereof. Background technique [0002] With the popularization of mobile terminals and the acceleration of the Internet, short, flat, fast and high-traffic communication content has gradually gained the favor of major platforms, fans and capital. With the emergence of the Internet celebrity economy, a number of high-quality short video platforms such as Douyin, Miaopai, Kuaishou, etc. have gradually emerged, and many people have even become Internet celebrities because of the live video platform. Others share your little bits and pieces, share interesting stories about your life, and can also push products or help people bring goods in the video. The operator plays the video data provided by the multimedia content provider to the user through the network live broadcast platform. Objectively speaking, when a us...

AI Technical Summary

Problems solved by technology

[0003] However, in the current video data pre-roll advertisements, more text and voice information are displayed, and the visual effect is poor, so the following disadvantages occur:

[0004] 1. The object is single, and the recognition subject mostly uses the human face as the detection object: In the field of to c, the existing technology of patch implanting video mostly uses the human face as the recognition object, and the implanted patch is related to the human face, and cannot Accurately identify the specific product model and brand to be pushed;

[0005] 2. Single video patch: the existing technology implanted patch material is called single, the difference between users is small, it is not easy to distinguish, and it is not conducive to publicity;

[0006] 3. The effect of video clips is weak: most of the video clips are static pictures or dynamic pictures with weak dynamic effects, and the publicity effect is weak;

[0007] 4. The video sharing mechanism is simple: there is little incentive mechanism for sharing, it cannot be found continuously, and it is difficult to form a virtuous circle;

[0008] 5. Low commercial value: It is impossible to form a fast and effective product promotion closed loop, and commercial transformation is difficult;

[0009] 6. The quality of short video production is not uniform: small studios or individuals are not easy to control the production quality of short videos, resulting in damage to the brand image;

[0010] 7. The production efficiency of short videos is low: At present, advertising companies recruit a large number of tail Internet celebrities to shoot short videos, and distribute them to small studios for video post-processing (including advertising placement, etc.), which costs a lot and takes a long time

Images