Industrial cameras are an essential component of machine vision systems and play a crucial role in these systems. They have been widely adopted in numerous fields, including online inspection on industrial production lines, intelligent transportation, machine vision, scientific research, military science, aerospace, and more. The main parameters of industrial cameras include resolution, frame rate, pixel count, pixel size, and spectral response characteristics. Next, we’ll delve into the details about the frame rate of industrial cameras:

Frame rate is a metric used to measure the number of frames displayed. The unit of measurement is frames per second, abbreviated as FPS or “hertz” (Hz).

Due to the unique physiological structure of the human eye, when the frame rate of a displayed image exceeds 16 frames per second (fps), it appears continuous—a phenomenon known as persistence of vision. This is precisely why movie film is shot frame by frame and then played back at high speed.

The frames per second (fps), or frame rate, refers to the number of times a graphics processor can update the image per second when rendering scenes. A higher frame rate results in smoother and more realistic animations. Generally, 30 fps is considered acceptable; however, boosting performance to 60 fps can significantly enhance interactivity and realism. That said, beyond about 75 fps, the improvement in smoothness typically becomes difficult to perceive. If the frame rate exceeds the monitor’s refresh rate, it will only waste the graphics processing power, since the monitor cannot update at that speed. As a result, any frames generated above the refresh rate are essentially wasted.

Maximum Frame Rate/Line Rate: This refers to the rate at which the camera acquires and transmits images. For area-array cameras, it is typically the number of frames captured per second (Frames/Sec.), while for line-scan cameras, it is the number of lines captured per second (Hz).

The relationship between camera frame rate and exposure time:

Someone asked: Why does the frame rate of an industrial camera drop—and drop significantly—when we increase the camera’s exposure time? What’s the relationship between the camera’s frame rate and its exposure? And if we want to maintain a fixed frame rate, how should we set the camera’s exposure time? To address these questions, I’ve written this article. I’ve also conducted relevant tests using a Sentech camera, and the results regarding the relationship between frame rate and exposure time are consistent with what’s described in this article. For a detailed explanation of the underlying principles, please see below.

Exposure and sensor readings

The image acquisition process in a camera consists of two distinct stages. The first stage is exposure. After exposure is complete, the second stage—the readout process—begins, during which data is read from the sensor’s registers and transmitted outward (the readout process).

In the image acquisition process, there are two common methods for camera operation:“non-overlapped”exposure and“overlapped”exposure. In non-overlapping(“non-overlapped”)In this mode, during each image acquisition cycle, the camera must complete the entire exposure/readout process before the next image acquisition begins.

Although non-overlapping(“non-overlapped”)This mode is suitable for many scenarios, but it’s not the most efficient approach. To increase the camera’s frame rate, the image data acquired in the previous frame can be read out and transmitted as soon as the next frame begins its exposure. The camera operates in “overlap” mode.(“overlapped”)The method of exposure.

From the figure, we can see that there is an overlap between the camera’s readout data and the start of exposure for the next frame. At the same moment, the camera performs two operations simultaneously, resulting in—within the same unit of time—in...In “overlapped” exposure mode,More images can be captured, meaning the camera has a higher frame rate.

From the two figures above, we can see that in...“non-overlapped”exposure and“overlapped”Under the exposure mode, there is a relationship among the periods of a single frame image as follows: In the "overlapped" exposure mode: FramePeriod ≤ Exposure Time + Readout Time

“non-overlapped”in exposure modeFrame Period > Exposure Time + Readout Time

Take STC-A202A as an example:

As indicated in the Spec, the pixel frequency is 36.8181 MHz; therefore, the duration of one clock cycle is 1/36.8181 MHz = 27.3836 ns. Next, let's examine the camera's timing chart. First, we'll look at the horizontal timing, as shown in the figure:

From the figure, we can read that 1 CLK = 27.1605 nanoseconds, which is roughly consistent with the pixel frequency we obtained from the specification. It takes 1920 CLK to scan one horizontal line; therefore, 1H = 27.1605 × 1920 = 52,148.16 ns = 52.14816 μs.

Next, let's look at the camera's Vertical Timing, as shown in the figure:

From the figure, the information we can read is that 1H = 52.1482 microseconds, and as we have determined through...FigureThe horizontal scanning time calculated is consistent. In one frame of image, 1252H need to be scanned, of which the number of effective pixels is 1220H. Thus, in one frame of image, the time required to read out the effective pixels is 1220 × 52.1482 = 63,620.804 μs = 63.620804 ms. For a single VD signal, the time required is: 1252 × 52.1482 = 67,793.5464 μs = 67.7935464 ms. According to our previous theory, the duration of one frame of image within a single cycle time is:Frame Period = Exposure Time + Readout Time

We know that the frame rate of STC-A202A is 15 fps, meaning that the frame period is 1/15 = 66.7 ms.

So in“non-overlapped”In exposure mode, the Exposure Time = Frame Period – Readout Time = 66.7 ms - 63.6 ms = 3.1 ms. In this mode, if the exposure time exceeds 3.1 ms, the frame rate will be lower than the standard frame rate of 15 fps.

Suppose a camera has a readout time of 66.7 ms, meaning the data transfer time is A = 66.7 ms, and the exposure time is B = 5 ms. Then the time required for one frame is C, where C = A + B = 66.7 ms + 5 ms = 71.7 ms. Consequently, the frame rate of this camera is 1000/71.7 = 13.94 FPS. Therefore, the frame rate of this camera is 13.94.

Note: In general, if our camera has a frame rate of 15 fps, it means the time required to read out the camera’s data is 1000 ms / 15 = 66.7 ms.