Smart Image Cropping Using Saliency Maps for Multi-Device Displays
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face challenges in automatically cropping images to ensure visually pleasing displays across devices with varying screen sizes, orientations, aspect ratios, and resolutions, as a single crop may not effectively include important image content without overlapping with display elements.
Innovation Solution
The system uses saliency maps and object detection to determine regions of interest (ROIs) within an image, calculating a cropping score based on the inclusion of essential and preferred content areas, and interpolating scores to recommend optimal crop dimensions and locations for different display configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a single crop is taken from an image for use across multiple devices, then the cropping process is simple and fast, but the image quality and visual appeal vary across devices with different display properties
Solution Approach 1:
The patent applies dynamics by making the cropping parameters adaptive rather than fixed. The system dynamically adjusts crop parameters (coordinates, dimensions, aspect ratio) based on the target display's specific properties such as screen size, orientation, and aspect ratio. This allows the same image to be optimally cropped for multiple different devices without manual intervention for each device.
Solution Approach 2:
The patent changes parameters by systematically varying crop parameters based on display properties. The system modifies crop coordinates, dimensions, and aspect ratios according to the target display's characteristics. This parameter adaptation ensures that each device receives a crop optimized for its specific display properties, resolving the contradiction between processing efficiency and cross-device quality.
2Manufacturing precision
If the crop encompasses more important parts of the image, then the image content quality improves, but the crop may overlap with display elements such as text, titles, clocks, and battery indicators
Solution Approach 1:
The patent applies preliminary action by pre-identifying safe zones or exclusion areas on the display where display elements (text, titles, clocks, battery indicators) are likely to appear. The cropping algorithm uses this预先 information to adjust crop parameters and avoid these problematic areas, ensuring that important image content is included without overlapping with display elements.
Solution Approach 2:
The patent applies local quality by treating different regions of the display differently. The system identifies areas of the display that are safe for image content versus areas that contain or may contain display elements. By applying different quality requirements to different local regions, the system maximizes important content inclusion while avoiding harmful overlaps with display elements.
3Manufacturing precision
If the crop is optimized for one device's display properties, then the visual appeal on that device is maximized, but the crop may not be suitable for other devices with different orientations and aspect ratios
Solution Approach 1:
The patent applies universality by creating a cropping system that serves multiple functions across different devices. The system generates crop parameters that are universally applicable to various display types, orientations, and aspect ratios. By designing the cropping algorithm to be device-agnostic and adaptable, a single cropping process can optimize images for multiple different devices, making the solution both precise for each device and versatile across devices.
4Manufacturing precision
If manual cropping is performed to achieve optimal results for each device, then image quality is maximized, but the time and complexity of the cropping process increases significantly
Solution Approach 1:
The patent applies self-service by enabling the cropping system to automatically determine optimal crop parameters without requiring manual intervention. The algorithm autonomously analyzes the image, identifies important content regions, and calculates appropriate crop parameters based on target display properties. This automation achieves high-quality results while eliminating the complexity and time requirements of manual cropping for each device.
Solution Approach 2:
The patent applies feedback by using the target display's properties as input feedback to the cropping algorithm. The system takes display characteristics (size, orientation, aspect ratio) as feedback and automatically adjusts crop parameters accordingly. This feedback mechanism enables the system to achieve optimized results for each device type without manual intervention, resolving the contradiction between quality and complexity.
Data Source
Figure 1A
Figure 1B
Figure 2A
AI summary
Devices, methods, and non-transitory program storage devices are disclosed to provide for automatic cropping of images, given a requested target dimensions and/or aspect ratio, e.g., by using saliency maps to identify the parts of the image containing the most important content-and ensuring that such content is, if possible, included in a determined cropped region from the image. In particular, the devices, methods, and non-transitory program storage devices disclosed herein may: define a first region of interest (ROI) in a given image that is most essential to include in an automatically-determined cropped region; define a second ROI in the given image that would be preferable to include in the automatically-determined cropped region; and then determine a cropped region from the given image, based on the requested target dimensions and/or aspect ratio, that attempts to maximize an amount of overlap between the determined cropped region and the first and/or second ROIs.