Learn about the application of machine vision sensors and how to select them

Machine vision sensor refers to an instrument that uses optical components and imaging devices to obtain image information of the external environment, and usually uses image resolution to describe the performance of the vision sensor. The accuracy of a vision sensor is not only related to the resolution, but also to the detection distance of the object being measured. The farther away the object is from each other, the worse its absolute position accuracy becomes.

Applications of machine vision sensors

1. In the car assembly plant, check whether the glue beads applied by the robot to the door frame are continuous and have the correct width. In the bottling plant, it is verified that the caps are properly sealed, that the fill level is correct, and that no foreign objects have fallen into the bottles before the caps are closed.

2. In the packaging production line, make sure to paste the correct packaging label in the right place.

3. In the pharmaceutical packaging production line, inspect the blister pack of aspirin tablets for broken or missing tablets.

4. In metal stamping companies, stamping parts are inspected at a speed of more than 150 pieces per minute, which is more than 13 times faster than manual inspection.

Tips for choosing a machine vision sensor

1. Determine the type of sensor according to the measurement object and measurement environment

In order to carry out a specific measurement work, it is necessary to first consider the principle of the sensor, which needs to be determined after analyzing many factors. Because, even if the same physical quantity is measured, there are multiple principles of sensors to choose from, which principle of sensor is more suitable, according to the characteristics of the measurement and the use of the sensor conditions need to consider the following specific issues: the size of the range; The requirements of the measured position for the volume of the sensor; whether the measurement method is contact or non-contact; Signal extraction method, wired or non-contact measurement; The source of the sensor, domestic or imported, whether the price is affordable, or self-developed.

After considering the above questions, it is possible to determine which type of sensor to use, and then consider the specific performance indicators of the sensor.

2. Selection of sensitivity

In general, within the linear range of the sensor, the higher the sensitivity of the sensor, the better. This is because only when the sensitivity is high, the value of the output signal corresponding to the change to be measured is relatively large, which is conducive to signal processing. However, it should be noted that the sensitivity of the sensor is high, and the external noise that is not related to the measurement is also easy to mix, and will also be amplified by the amplification system, affecting the measurement accuracy. Therefore, it is required that the sensor itself should have a high signal-to-noise ratio to minimize the factory disturbance signal introduced from the outside world.

The sensitivity of the sensor is directional. When the measured quantity is a single vector and its directionality is required to be high, the sensor with low sensitivity in other directions should be selected; If the measurement is a multi-dimensional vector, the smaller the cross-sensitivity of the sensor, the better.

3. Frequency response characteristics

The frequency response characteristics of the sensor determine the frequency range to be measured, and the measurement conditions must be maintained without distortion within the allowable frequency range, in fact, the response of the sensor always has a certain delay, and the shorter the delay time, the better.

The frequency response of the sensor is high, and the frequency range of the measurable signal is wide, while due to the influence of structural characteristics, the inertia of the mechanical system is larger, and the frequency of the measurable signal is lower because of the low frequency of the sensor.

In dynamic measurements, the response characteristics should be based on the characteristics of the signal (steady-state, transient, random, etc.) to avoid over-fire errors.

4. Linear range

The linear range of a sensor is the range in which the output is proportional to the input. Theoretically, the sensitivity remains constant within this range. The wider the linear range of the sensor, the greater the measuring range and the measurement accuracy. When selecting a sensor, when the type of sensor is determined, the first thing to see is whether its range meets the requirements.

But in reality, no sensor can guarantee absolute linearity, and its linearity is also relative. When the required measurement accuracy is relatively low, within a certain range, the sensor with small nonlinear error can be approximately regarded as linear, which will bring great convenience to the measurement.

5. Stability

The ability of a sensor to remain unchanged after a period of use is called stability. In addition to the structure of the sensor itself, the factors that affect the long-term stability of the sensor are mainly the environment in which the sensor is used. Therefore, in order for the sensor to have good stability, the sensor must have strong environmental adaptability.

6. Accuracy

Accuracy is an important performance index of the sensor, and it is an important link related to the measurement accuracy of the whole measurement system. The higher the accuracy of the sensor, the more expensive it is, so the accuracy of the sensor only needs to meet the accuracy requirements of the entire measurement system, and it is not necessary to choose too high. This makes it possible to choose between a number of sensors that meet the same measurement purpose that are relatively inexpensive and simple.

If the purpose of measurement is qualitative analysis, a sensor with high repeatability can be selected, and it is not appropriate to choose a sensor with high absolute accuracy; If precise measured values are necessary for quantitative analysis, sensors with the appropriate accuracy class are required.

For some special use occasions, if a suitable sensor cannot be selected, it is necessary to design and manufacture the sensor by yourself. The performance of the self-made sensor should meet the requirements of use.