日本語 alt English
CMICRO Corporation
Image sensors

Q. What is the difference between the CCD sensor and the CMOS sensor?

  "CCD" and "CMOS" are semiconductor devices with properties called photoelectric conversion that convert light into electrical signals. Both CCD and CMOS use photodiodes to convert incident light to electric charge and the accumulated charge is read out.
  CCD was the first to become popular. In CCD, the charge stored in each pixel is transferred to an output circuit, like a bucket relay, using a transfer CCD. The charge is then converted to a voltage at the output unit. On the other hand, in CMOS, each pixel independently amplifies the charge and has a voltage conversion circuit.
  Because CMOS sensors can share processes with general-purpose semiconductor devices and the cost can be reduced, most of the area sensors used in consumer cameras are now CMOS. Although CMOS sensors have been widely used to satisfy the needs for higher scanning speed, CCD sensors are still used in certain fields where high image quality is required.

Q. What is the line sensor?

  The principles and concepts of the line sensor are the same as those of the data reading sensor of a facsimile or copier. The line sensor is a sensor with image pickup devices (pixels) which are arranged as a line with high density. The number of pixels of cameras from CMICRO Corporation ranges from 512 to 16,384. There are two types of sensors: monochrome type sensors which are mainly used for inspecting plain test objects, and color type sensors which are used for testing the color such as printing tests. Recently, it has become to use line sensors which can detect near-infrared light.

Q. What is a multi-line sensor?

  Multi-line sensor means a line sensor with multiple lines.
  For example, a sensor that adds two lines of pixel output together produces twice the sensitivity of each pixel.
  On the other hand, there is a method called "Time Delay Integration," in which two or more lines up to 128 lines (number of steps) are used to forward the charge in the flow direction and the accumulated charge for the number of steps are outputted. The pixel's sensitivity and S/N are increased according to the number of steps and the pixel size is not changed. Therefore, the resolution in the flow direction is not changed. However, the charge transfer must be synchronized with the speed of the flow direction of test objects.

Q. What is binning?

  Binning is a method which two adjacent pixels of the sensor are regared to be one pixel and the sensitivity is increased. The number of pixels becomes 1/2, but the shooting time is not increased and the sensitivity can be doubled. The binning in the scanning direction is called horizontal binning, and the binning in the flow direction is called vertical binning.

Inspections using cameras

Q. What is the difference between the area camera and the line scan camera?

  Area cameras use area assemblies in which imaging devices are arranged in vertical and horizontal planes to obtain two-dimensional rectangular images. In other words, all consumer cameras use area sensors.
  On the other hand, line scan cameras are cameras that use line sensors to continuously capture images captured as lines. The following shows the image pictures of the area camera and line scan camera. By using line scan cameras, you can obtain very high-resolution images utilizing the larger number of pixels in the line. Furthermore, it is a very suitable camera for inspecting the high-speed, continuous flow of test objects without necessarily requiring joint processing of frames generated by area cameras. Line scan cameras are also called line cameras or line sensor cameras.

Q. In what fields are line scan cameras used?

  It is mainly used to inspect materials and processed products for industrial and food products. Industrial products that can be developed in two dimensions, such as sheets or rolls, are subject to inspection. They are used to detect damage and pinholes in packaging films, PETs, mirrors, glass for LCD TVs, wallpaper, and other products used in everyday life.
  Meanwhile, they are also used to inspect foodstuffs such as rice, cereals, tea leaves, seaweed, nuts, other ingredients, and processed dry fruits and cut vegetables as well.

Q. How is a color line scan camera realized?

  There are two main ways to achieve a color line scan camera.
  One is a trilinear method in which three line sensors are prepared corresponding to three colors of R, G, and B, and each sensor is equipped with a filter that passes only light of the wavelengths of R, G, and B. While this method is relatively inexpensive to manufacture, it is necessary to delay the output of each color line in synchronization with the movement of test objects to be inspected in order to correct the displacement of the three lines. In addition to the trilinear method, there are other methods utilizing multiple lines such as two-line method where a filter called a Bayer array is mounted on two lines.
  Another method is to use dichroic prism, which is described in the following section "What is a prism camera?"

Q. What is a prism camera?

  This is a camera that uses a prism to split the input light into different wavelengths such as R, G, and B. The figure below shows an example of a camera with a prism that splits the input light into three colors: R, G, and B.
  One line of input light is split into three bands of R, G, and B using a dichromic prism which transmits only light in a specific wavelength band and reflects light in the remaining wavelength band. The spectral light is received by sensors affixed to the output surface of the prism. There is no color drift between the 3 colors of R, G, and B, because the same line of input light is diffracted. It is possible to obtain a very high-quality image. However, because of the use of prisms, the camera is generally expensive and large in size.

Q. What kind of tests are near-infrared rays used for?

  Light at wavelengths between 780 nm and 2500 nm which is longer than visible light (380 nm to 780 nm) is called near-infrared light. Near-infrared light can transmit substances that are not transmitted by visible light. Light at specific wavelengths is absorbed by substances and the absorbed wavelength depends on the substance. This property is used to detect substances that cannot be distinguished by visible light.

Q. What kind of lights do you use?

  In the past, fluorescent lamps and halogen lights were used. In recent years, the use of LEDs for lighting has become more common due to the increasing brightness and long life, the increasing types of wavelengths that can be emited, and the cost reduction.

Q. What are the features of the CoaxPress interface?

  The CoaXPress interface is a communication standard between a camera and a frame grabber that enables high-speed video transfer of 6.25 Gbps (for Version 1.1) on a single coaxial cable, communication control, and power supply. This standard was proposed by the Japan Industrial Imaging Association in 2009 and was approved as an international standard in 2011.
  The cable length of 170 m (When the transmission speed is 1.25 Gbps) can be realized using coaxial cables which enable flexible connection configuration of connections. Additionally, the higher video transfer rate is achieved by using multiple cables. For example the video transfer rate becomes 12.5 Gbps by using two coaxial cables.

Q. How does a 3D line profile camera measure height?

  3D line profile camera measures the height by using a method called light-section method, in which a line-shaped laser beam is illuminated on a test object and the height is measured at the position of the sensor where the reflected light is received. A simple principle diagram is shown below.