Why are there more and more mobile phone cameras? What exactly do they do?

In the past, there was only a quiet camera behind the phone, and it didn't play at all. And now the mobile phone, two cakes and three tubes? Even? Yuba? The stream is endless.

What exactly do these cameras do?

To explain the reason of three shots, we have to start with two shots. Although there are many kinds of mobile phones supporting dual cameras on the market, most of them are equipped with a main camera and an auxiliary camera. According to the focus of the photo, there are generally four ways to match:

Color camera+black and white camera

Color camera+color (depth of field) camera

Color (wide angle) camera+color (telephoto) camera

Color camera+depth camera

The first three are the most common collocation methods on mobile phones at present. The essence of these three imaging methods is to process and superimpose the pictures taken by two cameras, so as to achieve different imaging effects. For the accuracy of image processing, the difference between the two images should not be too big, so the two cameras of all dual-camera phones are very close.

Color camera+black and white camera: to achieve clearer night photos.

Because the light is too dark at night, the photos we take are often very dark and we can't see anything clearly. In order to make the photos taken at night clearer, it is necessary to increase the amount of light. There are three ways to increase the amount of light: increasing the aperture, prolonging the exposure time and improving the ISO sensitivity. At present, the aperture of most mobile phones is not adjustable. Prolonging the exposure time can increase the amount of light, but it will bring the problem of jitter. Although most mobile phone manufacturers are equipped with optical anti-shake function at present, the ability of optical anti-shake is limited after all.

Digital camera is essentially photoelectric conversion, which improves the sensitivity, which is equivalent to multiplying the optical signal by the electrical signal. Electronic noise may also be recorded as a normal signal, which will increase the noise in the photo and make the photo unclear.

High noise

When the method of increasing the light input by materiality encounters a bottleneck on the mobile phone, it is necessary to invite the power of the algorithm.

First of all, here is a concept: Bayer filter. At present, most image sensors of digital cameras, scanners and other equipment use it to make color images. When we take pictures with a color camera, the natural light will be filtered by Bayer filter, and after a series of complicated arrangements and algorithms, a color image will finally be generated. In this process, the light is lossy. Black-and-white camera is not equipped with Bayer filter, which will let all light pass through, get higher light input, make the captured image clearer and keep more details.

Under the condition of night scene, better photos can be obtained by fusing clear black-and-white images with color images.

Color camera+color (depth of field) camera, bringing beautiful background blur.

What is the depth of field? When we take pictures, there is a clear range of scenery before and after the object (focus), and this range is the depth of field. When shooting portraits or close-ups, SLR blurs the background and highlights the subject by adjusting the depth of field.

Background fuzzy close-up

SLR is realized by adjusting the aperture of the lens. As we said before, the aperture of most mobile phones is not adjustable. So I want to take a close-up of a portrait with a blurred background, and my mobile phone has encountered a bottleneck, so we invite the power of the algorithm again.

First, let's do an experiment, close one eye, hold a pen in each hand, keep the two pens level with the eyes, and then touch the tips of the two pens. Have you ever found that two pen tips often don't touch each other, but if both eyes are open, the pen tips can always touch each other accurately. This experiment adopts triangulation method. Triangulation is to use two or more detectors to detect the target in different directions, and then determine the position and distance of the target by using the principle of trigonometric geometry.

Triangular method

The blur effect of dual-camera mobile phone simulating SLR is the triangulation method. Firstly, two cameras are used to determine the spatial distance of the object in the photo, that is, the depth of field information; Then the object in the photo is divided into planes according to the depth of field information; Finally, select the corresponding plane according to the requirements, for example, in portrait mode, select the plane where the portrait is located, and then blur other planes to get a photo with clear portrait but blurred background. Of course, this blurring process also needs strong algorithm guarantee, so I won't go into details here.

Color (wide-angle) cameras and color (telephoto) cameras provide better zoom capabilities.

What are wide angle and telephoto? Wide-angle lens has short focal length and large viewing angle, which is suitable for shooting large-area scenes; The telephoto lens has a small viewing angle but a large focal length, which is suitable for shooting distant objects. Generally speaking, the wide-angle shooting range is large and the telephoto shooting distance is far.

wide-angle lens

The way to change the focal length of SLR lens is optical zoom, that is, the viewing angle and focal length are changed by moving the lens back and forth in the lens. In order to zoom in this way, it is necessary to leave room for the lens to move, so the zoom lens is usually very large. Limited by the size and thickness of mobile phones, optical zoom has encountered bottlenecks in mobile phones.

long shot

In the past, mobile phones could zoom, but it was digital zoom. The so-called digital zoom is actually only a part of the viewfinder. Now some mobile phones have added optical zoom, but it is difficult to achieve better optical zoom in the case of limited space, at least for mobile phones now.

This camera combination method can achieve better zoom effect and get clearer zoom photos by switching cameras and merging photos taken by two cameras. This method is still not a real optical zoom, but the zoom effect is much stronger than that of a single camera.

Color camera+depth camera: It can be used in face recognition, AR, three-dimensional space measurement and other fields.

Depth camera is a kind of camera that can measure the physical distance between the object and the lens. This combination method can obtain real three-dimensional images, which has a wide application space in the future, but it is rarely used on mobile phones at present, so I don't want to introduce it more.

So back to our original question, why are there more and more cameras on mobile phones?

The principle of "three shots and four shots" is the same as that of "two shots", but on the basis of "two shots", more cameras are added to complete a richer combination.

In addition, the cost of the camera itself is not high, and mobile phone manufacturers can create a feeling that the cost of taking pictures of mobile phones is lower through the combination of more cameras, which is very flagship and more conducive to the promotion and sales of mobile phones.

Can mobile phones completely replace SLR?

In recent years, the pixel of mobile phone is getting higher and higher, even reaching 100 million pixels, so many people say: mobile phone camera? 、? Hang a SLR? Wait a minute. So can mobile photography really replace SLR? Is the pixel as high as possible?

The answer is: at least not yet.

This will invite an important role: the image sensor. The image sensor is a kind of photosensitive element, which converts the optical signal received by the camera into an electrical signal. For a camera, it is equivalent to a human eye. Therefore, the quality of the image sensor can basically determine the quality of the photo. However, due to the current technology, the size of the image sensor on the mobile phone cannot be enlarged.

The image sensor basically determines the quality of the photo.

Usually, a camera's 20 million pixels means that there are 20 million photosensitive spots on the image sensor. When taking a photo, the photosensitive point records the brightness and color of the light to generate a pixel, and finally all the pixels form a photo together. The more pixels (photosites) there are, the clearer the enlarged photo will be.

It seems that the higher the pixel, the better, but not necessarily. For the same image sensor, we can divide it into 20 million photosites or 40 million photosites. In this way, although there are more photosensitive points, the area of each photosensitive point is smaller, so each photosensitive point can receive less light, which will affect the image quality presentation.

Therefore, the most important thing to take a good photo is the amount of optical signals that the image sensor can accept and process. There is a popular saying in photography circles: outsole killing, which means that the imaging quality of large-size image sensors should crush that of small-size image sensors. As for why the mobile phone can't replace the SLR, this picture of putting the image sensor P of the mobile phone on the SLR (the mobile phone image sensor in red circle) can be said to be very essence.