CMOS (Complementary Metal Oxide Semiconductor) photosensitive chips are the core components of camera modules, and their classification methods are diverse, mainly based on pixel structure, process technology, application scenarios, etc. The following are detailed classifications and characteristics:
—
### **1. Classified by pixel structure**
####* * (1) Front Side Illuminated (FSI)**
-* * Structural Features * *:
The light first passes through the metal circuit layer and then reaches the photodiode.
-* * Disadvantages * *:
The metal layer will block some of the light, resulting in lower sensitivity under low light conditions.
-* * Application * *: Early low-end cameras are gradually being phased out.
####* * (2) Back Side Illuminated (BSI)**
-* * Structural Features * *:
The photosensitive diode is located above the metal circuit layer, allowing light to directly enter the photosensitive area and reduce obstruction.
-Advantages:
Higher quantum efficiency (QE), good low light performance, suitable for high pixel miniaturization design.
-* * Application * *: Mainstream mobile phone cameras (such as Sony IMX series, Samsung ISOCELL).
####* * (3) Stacked BSI**
-* * Structural Features * *:
Stack the pixel layer (photosensitive area) and signal processing circuit layer (such as ISP) vertically instead of traditional parallel layout.
-Advantages:
-Further reduce chip area and enhance integration.
-Supports faster readout speeds (such as high-speed continuous shooting, HDR).
-* * Representative models * *: Sony Exmor RS (such as IMX989), Samsung ISOCELL HP series.
—
### **2. Classified by process technology**
####* * (1) Traditional CMOS**
-Using standard semiconductor technology, the dynamic range and noise control are relatively weak.
####(2) Global Shutter**
-* * Features * *:
All pixels are exposed simultaneously to avoid the jelly effect of rolling shutter.
-* * Disadvantages * *:
Low pixel density and high cost.
-* * Application * *: Industrial inspection, high-speed motion shooting (such as machine vision, drones).
####(3) Rolling Shutter**
-* * Features * *:
Line by line exposure, low cost but may result in dynamic distortion (such as tilting when shooting high-speed objects).
-* * Application * *: Consumer grade cameras (mobile phones, security, etc.).
—
### **3. Classified by Color Filter Array (CFA)**
####* * (1) Bayer Pattern**
-* * Structure * *:
RGGB (red green green blue) filtering arrangement is used to restore full-color images through interpolation algorithms.
-Advantages: Mature technology and low cost.
-* * Disadvantage * *: Loss of color information, requiring algorithm optimization (such as demosaicing).
####* * (2) Non Bayer array**
-Four Color Array (RYYB):
-Replace some green (G) with yellow (Y) to increase the light intake (such as the IMX600 series jointly developed by Huawei and Sony).
-Disadvantage: Requires complex algorithms to correct colors.
– **RGBW**：
-Increase white (W) pixels to improve sensitivity, but sacrifice color accuracy.
####* * (3) Monochrome/Black and White Sensor**
-Colorless filter with extremely high sensitivity, used for security night vision and industrial inspection.
—
### **4. Classified by pixel size**
|* * Type * * | * * Pixel Size * * | * * Features * * | * * Application Scenarios * *|
|—————-|—————|—————————–|—————————|
|Large pixel |>2.0 μ m | High sensitivity, low noise | High end mobile phone main camera, car camera|
|Conventional pixels | 1.0~2.0 μ m | Balanced resolution and photosensitive performance | Mainstream mobile phones and consumer electronics|
|Small pixel |<1.0 μ m | High resolution but poor low light performance | Ultra high pixel phone (such as 108MP)|
—
### **5. Classified by dynamic range technology**
####* * (1) Native High Dynamic Range (HDR)**
-* * Technology * *:
-Multi exposure synthesis: Short exposure captures bright areas, long exposure captures dark areas.
-Dual Gain: Combining high/low gain signals within the same frame (such as Sony DOL-HDR).
-* * Application * *: Car camera, security monitoring.
####* * (2) Single frame HDR**
-Realize single frame wide dynamic range through special pixel design (such as Split Pixel).
—
### **6. Classified by application scenario**
|* * Type * * | * * Features * * | * * Represents manufacturer/model * *|
|——————|——————————————-|——————————–|
|* * Mobile Camera * * | High Pixel, Small Size, BSI/Stackable | Sony IMX989, Samsung GN2|
|* * Security Monitoring * * | Strong Low Light Performance, Supports Infrared Sensing | Huawei Technology OV48C, Ansenmei AR0234|
|* * Car camera * * | High dynamic range (>120dB), high temperature resistance | Sony IMX490, Anson Mei AR0820AT|
|* * Industrial Inspection * * | Global Shutter, High Frame Rate, High Precision | Sony IMX253, Basler Ace Series|
|* * Medical Endoscope * * | Miniature, low-noise, supports 4K/8K | Toshiba TCD series, OmniVision medical sensors|
—
### **7. Other special types**
-Event based Vision:
Only responds to changes in brightness, with ultra-low latency (such as neuromorphic visual sensors).
-Quantum Dot Sensor (QDOT):
Using quantum dot materials to enhance color and sensitivity (laboratory stage).
—
###* * Summary**
The selection of CMOS photosensitive chips needs to be comprehensively considered:
-Sensitivity: BSI>FSI, large pixel>small pixel.
-* * Speed * *: Global shutter is suitable for high-speed scenes, and rolling shutter has low cost.
-Dynamic Range: HDR technology support is required for in car/security applications.
-* * Integration * *: Stacked design is suitable for devices with limited space, such as mobile phones.
Current technological trends: * * Stacked BSI+Small Pixel High Resolution+On Chip AI Processing * * (such as Samsung ISOCELL with AI).