計算攝影--google相機的王者之道

08-11

計算攝影--google相機的王者之道

來自專欄 All in Camera4 人贊了文章

本文系微信公眾號《大話成像》，知乎專欄《all in camera》原創文章

大話成像讀者QQ交流群：237427716 大話計算機視覺QQ交流群：651805282

本站教學視頻大話成像之《數字成像系統》32講已上線淘寶，搜索關鍵字『大話成像』即可找到，或使用鏈接

PC用戶： https://item.taobao.com/item.htm?id=568925736204：

手機用戶可使用鏈接： 複製這條信息￥adMo0tqbc2N￥後打開手淘

請購買了視頻的讀者加入QQ群：679050565，群名稱：數字成像系統課程，驗證消息請輸入購買課程時候的個人ID。

這幾天DXO的手機相機兵器譜實在是熱鬧，一個月之內，榜首三易其主。

先是iphone8 放了大招幹掉了google pixel，然後沒幾天，三星note8 又橫空出世來了個並列第一。

蘋果三星這兩個硬體巨人擠在鐵王座上還沒有把座位捂熱乎，軟體巨頭google的pixel 2又一下奪回了手機相機的王者榮耀。

Dxo的評價是：google pixel 2 創下了手機相機的新記錄

Test summary

The Google Pixel 2 not only takes the DxOMark Mobile scoring crown with a stunning Overall score of 98, but its video performance improves by 5 points over the original Pixel, putting it at 96 and at the top of our video ranking. The Pixel 2』s photo performance is really impressive at 99, and just one point short of the Samsung Galaxy Note 8, which scored an amazing 100 in our photo category.

google pixel 2的軟體ISP方案，居然在視頻質量上居然提升了5分，達到了96分，登上視頻質量的王座，這真是太強大了！！！不得不為google的camera 團隊點贊！！！

這是google 相機的基本特性：

1/2.6-inch 12MP sensor
f/1.8 aperture
Dual-Pixel Autofocus
Optical Image Stabilization
HDR+technology
軟體方案的人像模式

表面看起來沒有什麼特別，那麼google是如何打敗競爭對手的呢？

在產品發布會上，google的回答是：

Its a combination of the hardware with its computational intelligence.

Google 的camera 研發團隊的負責人Marc Levoy，以前是斯坦福大學計算機圖形實驗室的教授，在接受記者採訪時闡述了 google 的相機技術方向----

"I think the excitement is actually just starting in this area, as we move away from single-shot hardware-dominated photography to this new area of software-defined computational photography."

computational photography，計算攝影 , 就是google 相機的王者之道。

那麼計算攝影到底能做什麼呢？優勢在哪呢？

計算攝影是基於傳統數字成像而衍生的新的成像技術：包括四個主要部分：

Computational Processing，

Computational Imaging/Optics ,

Computational Sensor ,

Computational Illumination.

Computational Processing包含以下子技術：

Mosaicing,

Matting,

Super-Resolution,

Multi-Exposure HDR,

Light Field from

Mutiple View,

Structure from Motion,

Shape from X.

Computational Imaging/Optics 包含以下子技術：

Coded Aperture,

Optical Tomography,

Diaphanography,

SA Microscopy,

Integral Imaging,

Assorted Pixels,

Catadioptric Imaging,

Holographic Imaging.

Computational Sensor 包含以下子技術：

Artificial Retina,

Retinex Sensors,

Adaptive Dynamic Range Sensors,

Edge Detect Chips,

Focus of Expansion Chips,

Motion Sensors.

Computational Illumination包含以下子技術：

Flash/no flash,

Lighting domes,

Multi-flash for depth edges,

Dual Photos,

Polynomial texture Maps,

4D light source

具體到google pixel 2 相機，我們能看到哪些計算攝影的應用呢？

1：快速對焦。

google可以利用digital refocusing 技術實現超快速對焦。

同樣是使用了dual pixel PDAF的硬體，google的對焦做的比別人更快更穩定，在DXO評測中拿到了96分的最高分。

2.光學、數字防抖+去模糊演算法

3.更好的HDR+

The traditional way to produce an HDR image is to bracket: you take the same image multiple times while exposing different parts of the scene, which lets you merge the shots together to create a final photograph where nothing is too blown-out or noisy. Googles method is very different — HDR+ also takes multiple images at once, but theyre all underexposed. This preserves highlights, but what about the noise in the shadows? Just leave it to math.

google的HDR方法是欠曝光HDR多幀融合，可以更大的避免抖動，降低圖像配準的難度，利用強大的去躁演算法去除低曝光帶來的雜訊，而不必妥協於曝光。

4.更好的低光照拍照性能

google also claims that, counterintuitively, underexposing each HDR shot actually frees the camera up to produce better low-light results. "Because we can denoise very well by taking multiple images and aligning them, we can afford to keep the colors saturated in low light," says Levoy. "Most other manufacturers dont trust their colors in low light, and so they desaturate, and youll see that very clearly on a lot of phones — the colors will be muted in low light, and our colors will not be as muted." But the aim isnt to get rid of noise entirely at the expense of detail; Levoy says "we like preserving texture, and were willing to accept a little bit of noise in order to preserve texture."

一般消費相機在低光依然希望能保持色彩飽和度，但是就容易出現明細的彩色雜訊，google的優秀降噪演算法可以避免這種窘境。

5.軟體方案的人像模式，這個功能有點打臉雙攝像頭或者深度感測器，google用一個攝像頭就實現背景虛化這種人像模式的拍攝，而不需要額外的硬體。

關於計算攝影，還有很多在成像科學上的應用，比如光場相機，編碼快門，反射分割，結構照明等等。

總而來說，google的軟體相機方案具有更加靈活，更低成本，更快實現產品迭代的優點，在目前的相機競爭中可以說獨樹一幟，而且這種軟體方案，讓安卓相機不必完全依賴高通的平台更新節奏，能夠最大的實現產品快速優化。

歡迎大家關注我們的公眾號大話成像和大話計算機視覺，並加入我們的兩個QQ交流群交流一些工作學習中的問題。