Image Signal Processor Auto Zoom Focus Depth Map
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems in imaging devices, such as cameras and smartphones, face challenges in simultaneously performing auto zoom and auto focus accurately, especially when the number of phase-difference detection pixels is reduced during zooming, making it difficult to achieve clear images with precise focus.
Innovation Solution
An image signal processor that calculates auto focus and auto zoom information using depth map data to set a zoom area based on specific ratios, allowing for simultaneous auto zoom and auto focus operations with a single user touch, thereby simplifying the process of capturing clear images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If zooming is performed manually by touching the screen, then the angle of view can be adjusted, but the number of phase-difference detection pixels is reduced in a zoomed state, making it difficult to accurately focus
Solution Approach 1:
The system performs preliminary calculations of outer points and area ratios using depth map information before executing the zoom operation. By pre-determining the zoom area and ratio based on the touched object's depth characteristics, the system ensures that focus information is captured and processed in advance, maintaining focus accuracy even when zooming reduces the number of available phase-difference detection pixels
Solution Approach 2:
The depth map serves as an intermediary data structure that bridges the gap between the touched object and the zoom/focus operations. By using depth information to calculate outer points and determine zoom areas, the system can perform accurate auto zoom and auto focus simultaneously without relying solely on phase-difference detection pixels, thus resolving the contradiction between zoom capability and focus accuracy
2Ease of operation
If auto zoom and auto focus are performed sequentially, then each operation can be completed step by step, but the process requires multiple user interactions and increases operation time
Solution Approach 1:
The system merges the auto zoom and auto focus operations into a single simultaneous process. By calculating both the zoom area (based on outer points and ratios) and focus information (using depth map) in parallel from the same depth map data and touch input, the system eliminates the need for sequential operations, reducing both user interaction steps and overall operation time while maintaining ease of use
3Area of moving object
If the number of phase-difference detection pixels is reduced during zooming, then the zoomed view can be achieved, but accurate focus becomes difficult to obtain
Solution Approach 1:
The depth map acts as an intermediary that enables focus detection independent of phase-difference detection pixel availability. By using depth information to determine the zoom area and calculate focus parameters, the system can maintain accurate focus detection even when the number of phase-difference detection pixels is reduced during zooming operations
Solution Approach 2:
The system changes the parameter basis for focus detection from relying on phase-difference detection pixels to using depth map information. By switching to depth-based focus calculation methods when zooming, the system maintains focus detection precision across different zoom levels without being constrained by the reduced number of phase-difference detection pixels
Data Source
AI summary
An image signal processor may be configured to process an image based on simultaneously performing auto zoom and auto focus. An image signal processor may be configured to receive a first image from an image sensor may include processing circuitry configured to determine auto focus information on an object in the first image using a depth map, set a first area based on calculating a first outer point and a second outer point associated with the object using the depth map, set a second area based on a particular ratio that is determined based on the first outer point and the second outer point, and determine auto zoom information based on calculating a zoom ratio corresponding to a size of the second area.