HyperOIS: A Comprehensive Optical Image Stabilization Solution for Modern Smartphones

固件 微控制器 人工智能 像素 校准 计算机科学 Boosting(机器学习) 图像质量 图像(数学) 计算机视觉 计算机硬件 实时计算 数学 统计
作者
Chang Liu,Chenhui Yin,Yang Han,Gang Wang,Zijing Mao,Yan Chen,Zhongxuan Tu,Heng Sun
出处
期刊:IEEE Transactions on Consumer Electronics [Institute of Electrical and Electronics Engineers]
卷期号:70 (1): 696-703
标识
DOI:10.1109/tce.2024.3369029
摘要

Optical Image Stabilization (OIS) has become popular in modern smartphones because it improves the imaging quality significantly by reducing the image blur caused by the natural hand tremors. In this paper, we introduce the novel OIS system - HyperOIS, which integrates the OIS controlling algorithm in sensor hub of the smartphone's system platform, instead of in the camera module's Micro Controller Unit (MCU) as firmware, thereby acquiring extra computational power for more advanced features, such as gyro prediction and adaptive gyro bias fitting. We also introduce the strategies for camera module calibration, with a novel calibration method in order to reduce the calibration time consumption, therefore boosting the production efficiency by 20%. Under different vibration frequencies, the performance of HyperOIS is tested with Stabilization Rate (SR) metrics and achieves impressive results of -34.37dB to -26.90dB. HyperOIS has already been implemented in a series of Xiaomi and Redmi models, and we have compared their OIS performances with various top-selling products, using independent OIS evaluation standard - CIPA. The reports show that the images taken by Xiaomi 14 and HyperOIS contain the least blur amount of $39.17{\mu }m$ (approx. 16 pixels) with a typical 4K resolution.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
刚刚
Hosea发布了新的文献求助10
1秒前
瑾予淇安关注了科研通微信公众号
3秒前
5秒前
木虫发布了新的文献求助10
5秒前
Non0发布了新的文献求助10
5秒前
CC完成签到,获得积分10
7秒前
Akim应助小懒猪采纳,获得10
9秒前
EZIO完成签到,获得积分10
10秒前
华仔应助ddizi采纳,获得30
10秒前
11秒前
11秒前
ewryetru完成签到,获得积分20
11秒前
yuxinzhao完成签到,获得积分20
12秒前
14秒前
领导范儿应助呆瓜采纳,获得10
15秒前
Sulin发布了新的文献求助10
15秒前
15秒前
15秒前
冰西瓜完成签到 ,获得积分10
15秒前
踏实晓筠发布了新的文献求助10
16秒前
19秒前
SY完成签到,获得积分10
20秒前
21秒前
21秒前
21秒前
渊崖曙春应助细心的微笑采纳,获得30
22秒前
22秒前
Akim应助PercyZzz采纳,获得10
23秒前
24秒前
24秒前
EZIO发布了新的文献求助10
26秒前
flasher22发布了新的文献求助30
26秒前
研友_89mo68发布了新的文献求助20
27秒前
28秒前
三国时代发布了新的文献求助10
28秒前
28秒前
打打应助Jeo采纳,获得30
29秒前
吃土豆长大的马铃薯完成签到,获得积分10
30秒前
30秒前
高分求助中
Production Logging: Theoretical and Interpretive Elements 2500
Healthcare Finance: Modern Financial Analysis for Accelerating Biomedical Innovation 2000
Agaricales of New Zealand 1: Pluteaceae - Entolomataceae 1040
Les Mantodea de Guyane Insecta, Polyneoptera 1000
지식생태학: 생태학, 죽은 지식을 깨우다 600
Crystal structures of UP2, UAs2, UAsS, and UAsSe in the pressure range up to 60 GPa 570
Mantodea of the World: Species Catalog Andrew M 500
热门求助领域 (近24小时)
化学 医学 材料科学 生物 工程类 有机化学 生物化学 纳米技术 内科学 物理 化学工程 计算机科学 复合材料 基因 遗传学 物理化学 催化作用 细胞生物学 免疫学 电极
热门帖子
关注 科研通微信公众号,转发送积分 3466417
求助须知:如何正确求助?哪些是违规求助? 3059200
关于积分的说明 9065226
捐赠科研通 2749643
什么是DOI,文献DOI怎么找? 1508690
科研通“疑难数据库(出版商)”最低求助积分说明 696996
邀请新用户注册赠送积分活动 696733