Classification of industrial cameras commonly used in machine vision systems
According to different photosensitive chips
We know that the sensor chip is the core component of the camera. At present, the camera commonly used sensor chip has CCD and CMOS:
1. CCD cameras, CCDs are called charge-coupled devices, and CCDs are actually just a way to organize the electrons out of an image semiconductor.
2. CMOS camera, CMOS is called "Complementary Metal Oxide Semiconductor", CMOS is actually just the technology of putting the transistor on the silicon block, there is no more meaning.
Although CCD stands for "charge-coupled device" and CMOS stands for "complementary metal-oxide semiconductor", neither CCD nor CMOS is useful for image sensing. The true sensor is called "image semiconductor," and CCD and CMOS sensors are actually used. It is the same type of sensor as "image semiconductor". The image semiconductor is a PN-bonded semiconductor and the photon-explosive junction that can convert light becomes a proportional amount of electrons. The number of electrons is calculated as the voltage of the signal, the more light enters the image semiconductor, the more electrons are produced, and the higher the voltage output from the sensor.
Because the human eye can see the target below 1Lux illumination (full moon night), the CCD sensor can usually see an illumination range of 0.1 to 3 Lux, which is 3 to 10 times that of the CMOS sensor, so the image quality of the current general CCD camera Better than CMOS camera.
CMOS can integrate photosensors, amplifiers, A/D converters, memories, digital signal processors, and computer interface control circuits on a single silicon wafer. It has a simple structure, many processing functions, fast speed, low power consumption, and low cost. Features. The CMOS camera has the problems of poor image quality, small size of the image sensing unit, and low filling rate. After 1989, the “active image sensor unit” structure appeared, which not only had address switches for photosensitive elements and image sensors, but also signals. The circuits such as amplification and processing improve the photoelectric sensitivity, reduce noise, and increase the dynamic range, making some parameters similar to those of CCD cameras, and they are superior to CCD in terms of function, power consumption, size, and price, and are gradually widely used. . CMOS sensors can be made very large and have the same sensitivity as CCD sensors, making them ideal for special applications. The CMOS sensor does not require complicated processing and directly converts the electrons generated by the image semiconductor into voltage signals. Therefore, it is very fast. This advantage makes the CMOS sensor very useful for high-frame cameras. The high frame rate can reach 400 to 100,000 frames/second.
Divided by output image signal format
Analog camera
The analog video camera outputs the signal in the form of a standard analog video signal and needs a dedicated image capture card to convert it into digital information that the computer can process. Analog cameras are generally used in the field of television cameras and monitoring. They have the characteristics of good versatility and low cost. However, they generally have low resolution and slow acquisition speed. They are also susceptible to noise interference in image transmission, resulting in degraded image quality. A machine vision system that does not require high image quality. Commonly used camera output signal formats are:
PAL (CCIR in black and white), Chinese TV standard, 625 lines, 50 lines
NTSC (black and white for EIA), Japanese TV standard, 525 lines, 60 fields
SECAM
S-VIDEO
Component transmission
Digital camera
The digital camera integrates the A/D conversion circuit internally and can directly convert the analog image signal into digital information. This not only effectively avoids the interference problem in the image transmission line, but also eliminates the constraints of the standard video signal format. The signal output uses a more high-speed and flexible digital signal transmission protocol, which can be made in various resolution forms, and the current situation in which digital cameras are flourishing. Common digital camera image output standards are:
IEEE1394
USB2.0
DCOM3
RS-644 LVDS
Channel Link LVDS
Camera Link LVDS
Gigabit
Divided by cell arrangement
Face camera
The area camera is our common form. The pixels are arranged in rows and columns. Each pixel corresponds to a pixel on the image. The resolution we usually refer to refers to the number of pixels. It should be pointed out that the color image in our computer is generally represented by three values of R, G, and B for each pixel, but our general color camera is not the case. The following are introduced separately:
The black-and-white camera, each pixel corresponds to a pixel, the pixel has relatively uniform sensitivity to various wavelengths of light, and only the gray value of each pixel is acquired.
The BAYER-transformed monolithic color camera is used. Each pixel of this camera actually corresponds to one of the R, G, and B pixels. The three R, G, and B pixels are arranged according to a certain rule. The R, G, and B primary colors of each pixel are calculated based on BAYER interpolation based on the tristimulus values of the pixel and surrounding points, so the images obtained by this camera are often not available. Very good color effect, especially the corresponding edge position will have more obvious color distortion and loss of detail. At present, our common color camera is generally in this form, and its price is similar to that of a black and white camera of the same grade.
3CCD color camera, this camera corresponds to each pixel with R, G, B three photosensitive elements, the use of splitter prism to refract the incident light to three CCD target surface, respectively, photoelectric conversion to get R, G, B three The color value (see the figure below). The quality of the image obtained by this camera is good, there is no problem of loss of detail, but due to the complex structure of the camera, it is generally more expensive. In addition, because this camera adopts the method of dichroic prism, the path of light reaching each CCD target surface is not the same, so the lens needs to be designed in order to achieve a better image effect, so the camera using 3CCD also needs Equipped with a dedicated lens.
Line camera
The line camera is a special form. The pixels are arranged in a one-dimensional line, that is, there is only one line of pixels, and only one line of image data can be acquired at a time, only when the camera and the object move in the longitudinal direction. To get the two-dimensional image we usually see. So in the machine vision system is generally used in the case of continuous movement of the measured object, especially suitable for high speed and resolution requirements.
The black-and-white camera is also the most commonly used line-array camera. Each pixel corresponds to a pixel. The grayscale image is acquired.
3Line color camera, the camera has R, G, B three rows of pixels, respectively, red, green, blue three wavelengths of light sensitive, so each pixel will correspond to R, G, B three channel values, Color image data is formed. It should be noted that since the three rows of R, G, B pixels collect information not at the same position at the same time (see figure), motion correction is required in actual applications to obtain the desired color image. The general camera has the function of performing motion correction processing internally, but it is prone to color distortion at the edge of the image with very high contrast.
The 3CCD color camera has the same principle as the area array 3CCD camera. The incident light is respectively refracted on three CCD target surfaces by using a dichroic prism, and the values of R, G, and B colors are obtained by photoelectric conversion. The image quality is good, but special lenses are required.
