Imaging Apparatus Phase Difference Detection with Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image recognition and focus detection systems face challenges in enhancing detection accuracy for subjects and focusing positions, particularly due to limitations in phase difference detection and exposure variations across different pixels.
Innovation Solution
An imaging apparatus with a first pixel generating a first signal and second and third pixels generating second and third signals for phase difference detection, combined with a machine learning model for inputting images and detecting subjects and focusing positions, allowing for independent readout rates and exposure amounts, and employing horizontal and vertical pooling for improved detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If phase difference detection is used for focus detection, then focusing speed is improved, but detection accuracy deteriorates under exposure variations
Solution Approach 1:
The imaging element is divided into distinct pixel groups: first pixels for visible light imaging and second pixels for infrared light phase difference detection. This segmentation allows each pixel type to be optimized for its specific function, with the visible light pixels providing high-quality reference images for accurate focus detection while the infrared pixels enable fast phase difference measurement.
Solution Approach 2:
The visible light reference image acts as an intermediary that mediates between the fast but less accurate phase difference detection and the need for high detection accuracy. By comparing phase difference information against the high-quality visible light reference image, the system achieves both speed and accuracy in focus detection.
2Adaptability or versatility
If independent readout rates are used for different pixels, then detection flexibility is improved, but device complexity increases
Solution Approach 1:
The readout control unit is designed with multi-functionality to handle different readout scenarios: it can read out first pixels and second pixels simultaneously at different rates, or sequentially, or in interleaved patterns. This universal readout controller manages multiple pixel types and timing configurations through a single integrated unit, avoiding the need for separate dedicated readout circuits for each pixel type.
3Measurement precision
If exposure amounts differ across pixels, then phase difference detection capability is improved, but image uniformity deteriorates
Solution Approach 1:
Different exposure amounts are applied to different pixel groups based on their specific detection requirements. The second pixels (infrared) use a first exposure amount optimized for phase difference detection, while the first pixels (visible light) use a second exposure amount optimized for generating high-quality reference images. This local quality differentiation allows each region to be optimized for its specific function while maintaining overall system performance.
Data Source
AI summary
There is provided a detection method used in an imaging apparatus including an imaging element that has a first pixel for generating a first signal and second and third pixels for generating second and third signals for detecting a phase difference, and a memory that stores a machine-trained model, the detection method including: an input step of inputting a second image based on the second signal, a third image based on the third signal, or a fourth image based on the second signal and the third signal to the model as an input image; a subject detection step of detecting a subject included in a first image generated by the first signal, through the model on which the input step is executed; and a focusing position detection step of detecting a focusing position with respect to the subject, through the model on which the input step is executed.