Weak current College】 【digital image acquisition of five key factors】.

<br> On the speed and resolution requirements driving the machine vision system for shift to digital image acquisition technology. When from the analog to the digital camera you need to consider learning the five key factors. <br> <br> On a test, .the image resolution of two factors: the field of demand and image processing system to address the minimum size. Of course, the lens, light, mechanical devices and other factors also play a role, however, if we will discuss the limitations to pixels, .we can avoid receiving these topics and focus camera features. <br> <br> A basic example, if a beverage packaging system must verify that the box is filled before the seal, the camera is needed for the above content imaging, to verify the .existence of the 24 only closures. This is very easy to understand, because the bottle and lid is one-to-one, in this image the smallest feature is the cover, the cover must be solved. <br> <br> Once you .identify good color and minimal characteristics, requirements of camera resolution can initially determined. You can estimate when the box is imaging, Cap in a black background to prominence. With the closure variable round, image as the two edges of the circle formed by the interval .. Edge is where images from dark to light, or the place from bright to dark places these points. Spacing is the distance between the directly. <br> <br> At this point, there is a need to define each edge point represents the .number of pixels. In this application, each edge point consists of three pixels, space consists of four pixels are sufficient. Therefore, in the image by 25mm Cap 10 pixels to define. From here, we can determine if a pixel represents objects 2.5mm .. <br> <br> Now we can determine the resolution of the camera. Select the object to represent the camera 400mm horizontal resolution, the camera needs a horizontal resolution of minimum 400/2.5 = 160 pixels. In the vertical direction, camera need .250/2.5 = 100 pixels of vertical resolution. In addition to horizontal and vertical resolution each added 10% of the wealthy, because depending on the location of the venue object, this calculation off the smallest camera resolution approximately 176 (H) x 110 ( .V). <br> <br> If such a resolution of the camera available, so you can get the image shown in Figure 1. Note than a lighter background relative pixel represents closures. cluster In this picture, basically to be able to distinguish .here 24 bottle cap, but small is not too easy. <br> <br> If you select a higher resolution camera, such as 640 × 480, this is a very used, will produce higher quality images, it can get more image details .(such as mark), as shown in Figure 2. <br> <br> 1. improve the resolution of the advantages and disadvantages <br> <br> Although higher resolution camera helps improve accuracy by analysis of clearer and more precise image, .but reduced speed. Digital camera to transmit image data is represented by a series of figures representing the pixel value. A resolution of 200 x 100 pixels camera with 20000, 20000 numeric values are sent to the acquisition system. If the camera work in 25MHz data .rate, it transmits every 40 nanoseconds. This creates a whole image requires approximately 0.0008 seconds, corresponding to 1250 FPS. <br> <br> Set the resolution to 640 × 480 pixels will have 307200 is about 15 times above. Use the same 25MHz .data rate, collecting the entire image need 0.012288 seconds, or the equivalent of 81.4 frames / sec. These values are expected, actual camera frame rate will be lower, because we had to add exposure and adjust the number of times, but the camera resolution .increase would lead to the camera frame rate proportional decline this is obvious. Although various camera output configuration will not sacrifice the frame rate of increase of camera resolution, but it is also necessary to increase the complexity and higher costs. <br> <br> .2. speed and exposure <br> <br> In the choose a digital camera, the body imaging speed must take full account of the good. In the previous example, suppose that objects in exposure <br> <br> Does not move; .therefore available is relatively simple and cheap camera. This situation does not always exist, but would regularly encounter. In some application objects continuously move, in other words, they are just in a very short time is static. <br> <br> Stationary .or slow-moving objects: array camera best suited to a still or moving slowly object Imaging. Because the entire array range must once exposure, exposure time, any movement will cause the image blur. However, the motion blur can reduce exposure time or using .Flash to control. <br> <br> Fast moving objects: when moving object using a planar array camera, you need to consider the exposure time of which is in the camera, the number of moving objects, you also need to consider the object .with a minimum of pixel representation, which is the object resolution. In acquisition of moving object the image thumb rule is that the exposure must occur in the collection object move less than one pixel of the time. <br> <br> If your collection .of objects is to 1 cm / sec speed uniform movement, and object resolution has been set to 1pixel/mm, the maximum exposure time is 1 / 10 sec. Because the objects move a distance that is equal to one of the camera sensor pixels, .when you use the maximum exposure time is here will have a certain amount of blur. In this case, the general trend in the exposure time is set higher than the maximum speed, such as 1 / 20 sec, you can keep objects in mobile imaging .within half a pixel. If the same objects to 1 cm / sec speed mobile, object resolution 1pixel/Micron, so a second in the required maximum exposure is 1 / 10000. the exposure settings for faster depending on the camera, and you are able .to give enough light object to get a good image. <br> <br> In attempts to gain access to move objects with short exposure times when there is aSome additional techniques you can use. In this case, the need for a very short exposure .time of the camera place without this ability, that you can use the electronic shutter or Flash. If the application needs faster than a single output camera capabilities, consider configuring multiple output of the camera. <br> <br> 3. the frame rate .<br> <br> The camera's frame rate of the camera in a predetermined period is sent to a collection system, full-frame number, it is often also referred to as the number of frames per second. For example, resolution of 640 × .480 for a camera in the highest frame rate of 50 frames per second. Therefore, the camera needs in one exposure after 20 milliseconds to send the frame. Some cameras are read in the current exposure when could not be applied subsequently exposure, they require no .images a fixed time interval. And some type of camera can read an image at the same time collecting the next exposure. Therefore, in the moving body Imaging must be considered when reading time and camera. When there is no exposure time should further consider the .interframe time interval. <br> <br> 4. the frequency response and sensitivity <br> <br> All digital cameras are configured on the light energy sensitive electronic sensor. Cameras on light energy-sensitive light energy wavelengths typically 400 nm to 1000 .nm. In addition, in less than 400 nm enters into the ultraviolet spectrum, while the other is greater than 1000 nm enters into the infrared. Here are some of the objects emit light of a specific wavelength to imaging needs, and the wavelength of the .camera features the definition also need to be defined. Filters also apply to block unwanted wavelengths of light, but still need to know the camera on demand of the wavelength of light to the very good response. <br> <br> The camera's sensitivity to .define how the camera on a fixed exposure sensitivity. The camera's sensitivity can also be defined as LUX or DN / (nJ / cm ^ 2). "LUX" is the image engineers common vocabulary, which is used to define a range of visible light .sensitivity of Photonic unit on, and DN / (nJ / cm ^ 2) is no radiation, confined to the responses of visible light. In General, the two forms are described the camera will react to light. DN / (nJ / cm ^ .2) expression of radiation that 1nJ/cm ^ 2 predict exposure will output DN pixel data (numbers, also became a gray value). Gain is another of the available features, it provides various levels of sensitivity. The camera's sensitivity to a defined gain .settings down description. However, be careful of the camera in high gain settings will have a high sensitivity, but the increased noise can lead to reduced dynamic range. <br> <br> 5. bit-depth <br> <br> Digital .camera produces digital data, or a pixel value. To become digital, this data will have a number of bits per pixel, called pixel bit depth. This depth is usually from 8 to 16 bit. In monochrome cameras, bit-depth definition from dark .to light up the number of gray levels, where 0 represents completely dark, 255 (8-bit depth), speaking on behalf of a fully lit. From 0 to 255 value represents shades, close to 0 to 255 in dark gray, close to .almost white. 10-bit data 1024 grayscale level, 12-bit data provides 4096 shades of gray levels. <br> <br> Each application should carefully to determine the use of fine grey-scale level, or use a coarse gray. Machine .vision systems typically use 8 bit pixel, 10 bit or 12-bit means that makes data volume times, as transferring data of another type. This causes the system to reduce speed, because on the use of two types of pixel bits, but not all .bits are meaningful. Relatively high depth also increased the complexity of system integration, because the more high bit depth requires more bandwidth, especially if the camera has multiple output. <br> <br> The five key factors determine a machine vision system for image .acquisition, and also restricting the entire machine vision system for positioning and subsequent analysis of the image. Application for yourself, give full consideration to the five elements, image acquisition, make the right decisions, contribute to the hardware selection and cost control. Should learn .this experience, try to make your own detours, maximum to meet customer needs. <br> <br> <br> <br> <br>.