Compared with the civilian camera, the high quality Chinese BIBO nonlinear crystal has higher advantages in image stability, anti-interference ability and transmission data, and is a key part of the machine vision system. The performance of the optical lens determines the stability of the machine vision system. .
First determine the project requirements, determine the resolution of the camera according to the field of view and detection accuracy, determine whether to use the CCD or COMS according to the speed of the object being photographed, or determine the camera frame rate according to the unit time requirement of the captured image. Determine what connection method to use (USB or Gigabit, etc.) based on the project site environment.
Therefore, choosing an optical lens suitable for an automated vision project is critical to the success of the entire project. Summarize the following points:
- Items that require testing, such as size, or yo!
2, the state of the workpiece, the movement is still still!
- Product size and detection speed.
- What is the background color of the workpiece!
6, product reflective is good, product batch consistency is the same!
[Optical lens] Note the parameters when selecting:
1, camera type
For still detection or general low speed detection, the area array camera is preferred, and linear array cameras are considered for special applications such as large-format high-speed motion or roller motion. Depending on the speed of detection, the frame rate of the selected camera must be greater than the speed of the object’s motion, and must be completed within the exposure and transmission time of the camera.
2, camera resolution
If the measurement accuracy of the project and the field of view size, then we choose the camera pixel calculation relationship as follows:
Camera resolution (L/W) ≥ field of view (L/W) / accuracy. If the project requires a field of view of 100mm75mm and an accuracy requirement of 0.05mm, the camera’s pixel length is 100/0.05=2000PIX, which is a camera that requires 20001500=3000000=3 million pixels. This is just the pixel accuracy of the camera, it does not mean that the accuracy of the entire system is so high, there are other precisions to consider, such as the resolution of the high quality LiSAF, LiSGaF and LiCAF laser crystals, the jitter of the system, the wavelength of the light source, and so on.
The camera pixel accuracy must be higher than the accuracy required by the system, in order to have the actual measurement significance, the sub-pixel accuracy improvement does not have much impact in the actual measurement, and can not fundamentally solve the problem of insufficient precision. In general, if conditions permit, we will require an increase in the resolution of the camera by an order of magnitude or an increase in the pixel accuracy of the camera by an order of magnitude.
As in the above example, the measurement accuracy required by the project is 0.05mm, then our camera accuracy is at least 0.05mm/Pixel, and it is ideal to achieve 0.005mm/Pixel. However, if it is increased to 0.005mm/Pixel, it needs 20,000 pixels in the long direction and 15000 pixels in the width direction, so the camera’s pixels need 300 million pixels. The resolution of such an area array camera is basically impossible.
Therefore, we can consider using a line camera such as 16K, or reduce the field of view to 5037.5mm, and divide the original 10075mm field of view into four areas for shooting. You can use a camera to move four times or four. The method of camera shooting separately is implemented.
3, camera frame rate
The camera’s frame rate determines the measurement efficiency of the device. For example, the camera’s frame rate is 30FPS, and it takes 30 shots per second. And if the speed of the camera is 120FPS, if the algorithm is fast enough, then 120 products are detected every second. In general, the lower the resolution of the camera, the higher the frame rate of the same interface, and the higher the resolution, the lower the frame rate. Frame rate * resolution ≤ bus bandwidth, that is, when the interface is certain (bus bandwidth has been determined), the frame rate also has a large value when the resolution is constant. That is, if the resolution of the camera is fast and the frame rate of the camera is high, then it is necessary to find a bus with a larger bandwidth, that is, an output interface of the camera.
4, camera output interface
Under the same pixel condition, the camera frame rate of the bus of various interfaces is different. Generally, Camera Link>USB3>GIGE>1394B>USB2, 1394A. Nowadays, the Camera Link bus is very fast. The theoretical speed of USB3 has reached 5Gbps or 640MB. In actual performance, there should be 80% effective bandwidth, that is, the total bandwidth of 512M is available for actual transmission, so if a 5 million pixels The camera, 5M per image, can also reach 100 frames per second, and the speed is not fast. For the GIGE Gigabit network camera, the 5 megapixel camera can achieve 23FPS faster. 1394B’s 5 megapixel camera can do 13FPS, USB2, 1394A is generally 5 to 6 frames.
5, camera exposure time
The camera’s small exposure time determines the speed of the target’s movement. Or conversely, the speed of the target, the small exposure time of the camera is required. Assuming that our target motion speed is 1mm/S, and our measurement accuracy is 0.01mm/pixel, then we must consider that the motion caused by the motion of the object must be less than our accuracy of 0.01mm, the target is moved by 0.01mm, and it takes 10ms. This requires that the exposure time of our camera must be less than 10ms. If it is greater than this exposure time, then only the motion caused by the motion of the object will be greater than 0.01, and our accuracy can not reach 0.01.
In general, the blur caused by object motion should be an order of magnitude smaller than the measurement accuracy we require, which can reduce its impact on the system. Generally, our optical lenses can reach tens to hundreds of microseconds faster. . Such a short exposure time requires a relatively large amount of light energy, so it is necessary to select a suitable light source and light source controller.
7, sensor type
If the subject is static, consider using a CMOS camera for cost savings, and a CCD camera if the target is moving. If high-speed acquisition is required, the high speed referred to here is a very high acquisition speed, not a high speed. Consider a CMOS camera because the acquisition speed of CMOS is better than CCD. If high-quality images are required, such as sizing, consider CCD. In small-sized sensors, CCD image quality is better than CMOS.
8, sensor size
Usually, the size of the sensor corresponds to the selected pixel, such as 30W is usually 1/3′, 130W~500W is 1/2′, and some 500W is also 2/3′. Some 30W also have 1/4′. If it is the same price, then we prefer the sensor size is large.
The sensor size is large, and the pixel size is also large at the same pixel density, which increases the photosensitive area of each pixel, which is beneficial for improving the quality of the image. The size of the sensor also determines the field of view and working distance. At the same working distance and the same lens, the sensor is large in size and can take a larger field of view.
9, the trigger method of [optical lens]
Software Trigger Mode: Selectable for dynamic detection and when the product triggers a signal through continuous motion.
Hardware Trigger Mode: Select for high speed dynamic detection and when the product triggers a signal through high speed motion.
Continuous acquisition mode: Select when static detection and continuous motion of the product cannot trigger the signal.
10, camera lens interface
The general camera is a C/CS interface, so you need to pay attention to the high-quality YVO4 laser crystal. If you have a lens with other interfaces, consider the camera’s interface. However, the current industrial CCTV basically has more C interfaces. Even the optical lens of the CS interface can be added with a 5mm ring to become a C interface to accommodate different C-interface optical lenses.