Video Frame Cropping and Scaling for Fixed Target Object Size

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing systems struggle to maintain a consistent size for a target object in video frames as it moves relative to the camera, requiring manual adjustments and lacking intelligent tracking capabilities, especially in dual camera systems.

Innovation Solution

Implement systems and techniques that utilize object detection and tracking, combined with smoothing functions, to maintain the size and position of a target object across frames, using dual camera modes and lens switching algorithms for consistent object size and location.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If object detection and tracking is performed in video sequences, then motion information and behavior analysis are obtained, but target objects change size due to camera movement and perspective effects

Engineering Contradiction:
Improvetarget size consistencyVSAvoidtarget object scale
Core Design Contradiction:
Measurement precisionVSShape

Solution Approach 1:

The system dynamically adjusts the scale of target objects in output frames based on their detected positions and sizes in the video sequence. This allows the target objects to maintain a consistent, predetermined size in the output regardless of their varying sizes in the original video frames, effectively resolving the scale inconsistency caused by camera movement and perspective effects

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the size parameter of target objects in the output frames to a predetermined fixed size. By modifying this physical parameter, the system ensures that target objects maintain consistent dimensions across all output frames, eliminating the size variations that occur during video playback due to camera motion and perspective changes

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If video frames are processed to detect and track objects, then motion information is extracted, but processing time increases

Engineering Contradiction:
Improveobject detection accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs object detection and tracking in advance on the video sequence to obtain motion information, then uses this pre-processed information to generate the augmented output. This preliminary processing allows the actual output generation to be faster, as the complex detection and tracking operations are completed beforehand

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system creates simplified representations (bounding boxes) of detected objects and uses these copies for tracking and output generation. By working with these simplified copies rather than full image processing, the system reduces processing time while maintaining detection accuracy

Inventive Principle:
Principle #26Copying

3Stability of the object's composition

If target objects are resized to fixed dimensions, then consistency is improved, but original spatial relationships may be distorted

Engineering Contradiction:
Improvetarget size consistencyVSAvoidspatial relationship accuracy
Core Design Contradiction:
Stability of the object's compositionVSShape

Solution Approach 1:

The system resolves the spatial relationship issue by adding a new dimension - creating multiple output frames for each input frame. Each output frame contains a target object at the predetermined fixed size, and the collection of output frames collectively preserves the spatial relationships through temporal sequencing rather than spatial positioning

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Data Source

PatentEP4154511B1Maintaining fixed sizes for target objects in frames
Publication Date: 2026.05.06 QUALCOMM INC
  • EP4154511B1 patent drawingFigure 1
  • EP4154511B1 patent drawingFigure 2
  • EP4154511B1 patent drawingFigure 3

AI summary

Techniques are provided for processing one or more frames. For example, a region of interest can be determined in a first frame of a sequence of frames. The region of interest in the first frame includes an object having a size in the first frame. A portion of a second frame of the sequence of frames (occurring after the first frame in the sequence of frames) can be cropped and scaled to cause the object in the second frame to have a same size (and in some cases a same location) as the object in the first frame.