High-performance industrial camera

Industrial cameras are a key component in machine vision systems, and their basic function is to convert optical signals into ordered electrical signals with industrial-grade high quality. Industrial cameras generally consist of image sensors, internal processing circuits, data interfaces, IO interfaces, optical interfaces, etc.

The core of the camera is the sensor. Modern sensors are all solid-state electronic devices, each containing millions of discrete light-detecting areas called pixels.

I. Classification of Industrial Cameras

There are many types of industrial cameras. Figure 1 shows the classification of industrial cameras based on several different classification criteria.

图1 工业相机的分类

Sensor Type

According to the different types of camera sensors, they can be divided into two categories: CCD (Charged Coupled Device) and CMOS (Complementary Metal-Oxide Semiconductor).

A CCD sensor is a silicon chip containing numerous light-sensitive areas, as shown in the figure. The working principle of a CCD is to convert optical signals into electrical signals, which are then transmitted in sequence to a common output structure. The charge is then converted into voltage, and these signals are sent to a buffer and stored outside the chip, as illustrated in Figure 2.

Figure 2 Schematic Diagram of CCD Structure

In a CMOS sensor, the charge on the photosensitive pixels is converted into voltage within the pixel area. The signals are multiplexed by row and column and fed into the on-chip Digital-to-Analog Converter (DAC). Each area consists of a photodiode and three transistors, which perform the functions of pixel reset or activation, charge amplification and conversion, and selection or multiplexing, as shown in Figure 3.

Figure 3 Schematic Diagram of CMOS Structure

CCD and CMOS each have their own advantages in different application scenarios, as shown in Table 1.

Sensor	CCD (Charged Coupled Device)	CMOS (Complementary Metal-Oxide Semiconductor)
Design Technology	Single photosensitive device	Photosensitive device connected to an amplifier
Pixel Signal	Electronic charge packet	Voltage
Chip Signal	Analog	Digital
Sensitivity	Higher	Lower (due to small photosensitive aperture)
Fill Factor	High	Lower
Noise Ratio	Lower noise (single amplification)	Higher noise (millions of amplifications)
Power Consumption Ratio	Higher power consumption (requires external voltage)	Lower power consumption (direct amplification)
Information Reading Method	Complex (requires external voltage control)	Simple (directly reads current signals)
Information Reading Speed	Slower	Faster
Exposure Mode	Global shutter	Both rolling shutter and global shutter
Cost	Higher	Lower

Table 1 Comparison of CCD and CMOS Sensors

CCD has advantages such as high signal-to-noise ratio, strong transparency, and excellent color reproduction. It is widely used in high-end fields such as screen inspection, transportation, and medical care. However, the high cost and high power consumption of CCD have restricted the room for its market development. With the continuous improvement of CMOS technology and the decreasing price of high-end CMOS, CMOS will occupy an increasingly important position in the field of machine vision.

Shutter Type

A shutter is a mechanical device used to control the exposure time of a photosensitive element or film. To protect the photosensitive element or film in the camera from overexposure, the shutter is always closed. After setting the shutter speed, once the camera's shutter release button is pressed, the light passing through the lens will properly expose the photosensitive element or film during the time the shutter is open.

Sensors are divided into Global Shutter and Rolling Shutter according to different exposure forms. Generally speaking, CCD sensors all use global shutters, while CMOS sensors can use either global shutters or rolling shutters.

A global shutter means that all rows of pixels in the entire chip are exposed simultaneously, with each row of pixels starting and ending exposure at the same time. After exposure is completed, data begins to be read out row by row. The duration of exposure and data reading for the camera sensor is the same, but the timing of finishing data reading varies, as shown in Figure 4.

Figure 4 Principle of Global Shutter

A rolling shutter means that when the chip starts exposure, each row begins exposure sequentially in order. After the first row finishes exposure, it immediately starts reading out data; once the data is completely read, the next row begins to read out data, and this cycle continues. Pixels in different rows have different start and end times for exposure, as shown in Figure 5.

Figure 5 Principle of Rolling Shutter

When shooting moving objects, there are significant differences in the effects between global shutters and rolling shutters. If exposure is done using the global shutter method, all pixels are exposed within the same time period, which can "freeze" the object in the shot. In contrast, with the rolling shutter exposure method, since the start and end times of exposure for pixels in different rows vary, image distortion such as bending may occur. Therefore, rolling shutter cameras are usually applied in static or low-speed scenarios, while global shutters are typically used in dynamic situations, such as high-speed flying shots.

Data Interface

The data interface of an industrial camera refers to the interface through which the camera transmits data to a computer after image acquisition, with each interface following specific communication protocols. According to the type of data interface, it can be divided into various types of data transmission interfaces such as Gigabit Ethernet (GigE), 10 Gigabit Ethernet (10GigE), USB 3.0, Camera Link, and CoaXPress. Different data interfaces have different characteristics, and users can choose based on comparisons in aspects such as cable length, transmission speed, latency, and cost.

Interface	Gigabit	10 Gigabit	USB 3.0	CameraLink	CoaXPress
Speed/(Gbit/s)	0.1	1	0.3	0.64	2.56(4根)
Distance/m	100 (Twisted Pair) >100 (Fiber Optic)	100 (Twisted Pair) >100 (Fiber Optic)	5(Standard Passive Cable) >5(Fiber Optic)	10	45(CXP-6) 35(CXP-12)
Cost	Low	Middle	Low	High	High
Advantages	Good expandability High cost performance Good manageability and maintainability Good wide applicability	High bandwidth Good expandability Good manageability and maintainability Simple setup scheme	Supports hot swapping Easy to use Can connect multiple devices Cameras can be powered via cables	High data rate Strong anti-interference ability Low power consumption	Large data transmission volume Long transmission distance Selectable transmission distance and volume Low price and easy integration Supports hot swapping
Disadvantages	High CPU usage High requirements for host configuration	High CPU usage High requirements for host configuration	Poor stability Short distance	High price Requires separate power supply	High cost High complexity of setup

Table 2 Comparison of Data Interfaces

II. Performance Parameters of Industrial Cameras

Industrial cameras have many parameters, and customers can choose appropriate parameters according to their actual usage needs. The following will introduce the important performance parameters of the camera.

1、Resolution

Resolution is mainly used to describe a camera's ability to distinguish the object being photographed. When imaging in the same size of field of view, the higher the resolution, the more obvious the display of details. In general, the resolution of an area scan camera refers to the number of pixels (Pixels) of the camera's photosensitive chip. There are two forms of expressing camera resolution: one is expressed as "the number of pixels in the width direction of the photosensitive device × the number of pixels in the height direction of the photosensitive device", and the other is expressed as "the nominal resolution value of the camera", such as 2 million pixels.

2、Pixel Size

The pixel size refers to the actual physical size of each pixel on the pixel array of the chip. When the target size is the same, the smaller the pixel size, the higher the resolution.

3、Target size

The target size refers to the photosensitive area of the camera's image sensor, which is generally identified by the length of the diagonal, with the unit being in (inches). Under normal circumstances, when other conditions such as pixel size are fixed, the larger the target size, the higher the resolution of the camera.

4、Frame Rate/Line Frequency

Frame rate/line rate refers to the rate at which a camera captures and transmits images. For area scan cameras, it is generally expressed by the number of frames captured per second (Frames per Second, fps), i.e., frame rate. The higher the frame rate, the smoother the picture. For line scan cameras, it is generally expressed by the number of lines captured per second (Hz), i.e., line rate.

Frame rate (line rate) = Number of images (frames or lines) output by the camera per second / Time (s) consumed for outputting a single frame (line)

In general, the higher the resolution of the camera, the lower its frame rate. This is because a higher resolution camera captures larger digital image data, which may result in longer acquisition time and data transmission time. When the frame rate of the viewed picture is higher than 10-12 frames per second, the picture seen by the human eye is continuous. In industrial production, there may be situations where high-speed moving objects need to be photographed (commonly known as "flying shot"). In this case, this parameter needs to be considered emphatically, and a camera with a higher frame rate should be selected to avoid motion blur in the captured pictures.

5、Pixel Depth

Pixel bit depth (also known as pixel depth, image depth, or grayscale) refers to the number of bits used to store a single pixel. It determines the possible grayscale levels for each pixel in a grayscale image or the possible number of colors for each pixel in a color image. A higher bit count allows for more detailed readout, but a higher bit depth can slow down the frame rate.

6、Exposure time

Exposure time (shutter time) refers to the duration during which the pixels are exposed to light. Under the same external conditions, the longer the exposure time, the higher the brightness of the image. However, an excessively long exposure time will reduce the frame rate/line rate. Different cameras have different upper and lower limits of exposure. In some flying shot applications, insufficiently short exposure will cause image smearing, so industrial cameras need to have the capability of imaging within an extremely short exposure time.

7、Signal-to-Noise Ratio, SNR

The signal-to-noise ratio (SNR) refers to the ratio of signal to noise in an image, with the unit being decibels (dB). The higher the signal-to-noise ratio of an image, the better the image quality.

8、Dynamic range

Dynamic range refers to the difference between the minimum detectable light signal and the maximum detectable light signal, reflecting the range of light signals that an industrial camera can detect. The larger the dynamic range, the more obvious the display of details in the bright and dark areas of the image.

9、Spectral Range

The spectral response characteristic refers to the ability of a chip to respond to light of different wavelength bands, which is usually characterized by a spectral response curve. When selecting an industrial camera, it is generally necessary to choose a chip according to the wavelength band of the light source in the actual application scenario to achieve the best imaging effect.