Object-Aware Video Cropping for Automated Camera Motion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for achieving camera movement effects in videos require skilled manual operations during shooting or complex post-production editing, leading to high implementation costs and time-consuming processes.
Innovation Solution
An image processing method that identifies objects and their states in video frames to determine cropping positions and sizes, simulating lens movements without manual operations, using an image processing apparatus with modules for object identification, cropping, and scaling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If manual post-production editing is used to add camera movement effects, then camera movement effects can be achieved, but the operation process becomes complex and time-consuming
Solution Approach 1:
The system performs preliminary analysis of the video content to identify key objects and their states, then pre-determines the cropping positions and sizes needed to simulate camera movements. This preliminary processing enables automatic generation of camera movement effects without requiring time-consuming manual post-production editing, directly resolving the contradiction between automation and time consumption.
2Extent of automation
If skilled manual operations are used during shooting to achieve camera movement effects, then expressive video results can be obtained, but implementation costs increase due to skill requirements
Solution Approach 1:
The system enables videos to automatically generate camera movement effects through self-service processing. By analyzing video content and automatically determining cropping parameters based on identified objects and their states, the system eliminates the need for skilled manual operations during shooting, thereby reducing implementation costs while maintaining automation.
3Ease of operation
If manual editing operations are performed to create camera movement effects, then video expressiveness can be enhanced, but the operation process becomes cumbersome
Solution Approach 1:
The system replaces manual mechanical editing operations with an automated computational system. By using object identification and state analysis to automatically determine cropping positions and sizes, the system substitutes complex manual editing processes with automated image processing, thereby simplifying operations while reducing the perceived device complexity for the user.
Data Source
Figure 1
Figure 2a
Figure 2b
AI summary
Provided in the present disclosure are an image processing method, an apparatus, an electronic device and a storage medium. The method comprises: recognizing a target image frame to determine a first recognition object position and a second recognition object state; on the basis of the first recognition object position, determining a target cropping position and, on the basis of the second recognition object state, determining a target cropping size, the target cropping position being a cropping center position when the target image frame is cropped, and the target cropping size being a cropping size when the target image frame is cropped; and, according to the target cropping position and the target cropping size, cropping the target image frame and scaling same to a preset size. Executing the technical solution of the embodiments of the present disclosure can avoid complex manual camera moving processes, enrich picture infectivity, and improve user experience.