Tegra platform back camera driver implementation 2

Camera physical architecture

Generally, camera consists of two parts: lens and sensor IC. Some sensor ICPs integrate DSPs, while others do not integrate, but they also need external DSP processing. In terms of subdivision, the camera device consists of the following parts:

Lens

Lens is the lens. Generally, the lens structure of camera is composed of several lenses, which are divided into plastic lenses (plastic) and glass lenses (GLASS). Generally, the lenses are: 1 P, 2 p, 1g1p, 1g2p, 2g2p, 4G, etc.

Sensor

Sensor is an image sensor. sensor is a semiconductor chip and has two types: CCD and CMOS. Sensor converts the light uploaded from lens to an electrical signal, and then converts it to a digital signal through an internal adad. Since each pixel of the sensor can only light R, B, or G, each pixel is stored in a monochrome, which we call raw
Data. To restore the raw data of each pixel to three colors, the ISP must process the data.

ISP

The ISP is the image signal processing. It mainly completes digital image processing and converts the raw data collected by the sensor to the formats supported by the display.

In terms of hardware, the camera system is mainly divided into controllers and cameras. Its functions mainly include preview, takepicture photo and recording video.

IPU-Image Process Unit: The image processing unit used to control the camera and display.

Image Acquisition-image data information collected by camera is controlled through the CSI interface of IPU.

DMA ing to memory-IPU maps the collected data to a memory segment through DMA.

Queue mechanism-in order to transfer data more efficiently, the data in the memory is taken into one queue and transmitted to another queue.

Video output-extract the video data from the queue and use IPU to control the independent video storage. The video is finally displayed.