IDS Imaginghas released its new contrast-based autofocus features in the popular LE board level cameras. These additions take advantage of standard liquid lenses from Varioptic with resolutions up to 18 megapixels. The uEye software now comes with an intuitive GUI with adjustable regions of interest and various image sharpness measurement algorithms.
As much as “Auto focus” seems like it would be the flip of a switch, its important to understand the various methods used in the image analysis. In order to focus an image, algorithms are needed to measure image sharpness which is relayed to the liquid lens to make adjustments. These methods as based on principles in measuring edge sharpness to analyzing histogram values of the pixel grey scales.
Measuring image sharpness additionally has various algorithms which which can be run providing more exact methods versus basic analysis. It is important to understand these methods as additional processing power is required, effecting the overall camera frame rate.
IDS Imaging has a “Tech Tip”which covers various auto focus methods, defines the characteristics of search algorithms and how they effect speed and provides application examples. Click the icon below to download.
A great question and maybe you can, but what ARE the differences?
Basically, there are just a few questions you need to answer to see if you should use a webcam for you machine vision application which are as follows:
Do you need to program to integrate the video into an application with processing or control?
Do you need consistent image quality?
Are you doing computer vision (the computer is making decisions based on the images) or are you just viewing the images visually?
Do you care if the camera specifications change over your product’s life cycle?
Is the object under inspection moving?
Do you need to control when you take the picture or interface to a trigger or strobe?
Do you need to be able to choose what lens you will need?
If the answer to any of the above are YES, then a webcam will NOT work well or at all for your application. If the answers are NO, then by all means, you might be able to save money and just use a webcam. (You can stop reading here if you want, or continue for more details below).
Machine Vision Camera Software
Webcams do NOT come with a SDK as they are made to show video only. They normally provide a universal video driver, and also an application for viewing video.
Webcams have rolling shutter sensors which mean they cannot acquire images of moving objects without ‘smearing’ them. Industrial machine vision cameras use sensors with global shutters providing the ability to freeze the image to produce non smeared images of moving objects.
Trigger and Strobe Control
Webcams only have an interface to the USB data, whereas industrial machine vision cameras have hardware and software inputs and outputs. These allow for exact timing for a trigger to take a picture and a strobe to illuminate the object.
Camera Specs Changing over time
Webcams just need to show you video! In turn the manufacturers are not concerned if the sensors inside the camera change every six months. Whether the sensitivity changes by 10% makes no difference when you are just video conferencing with Grandma.
Industrial machine vision cameras are made with image sensors that don’t go obsolete every 6 months, but rather companies hope for 10 year life spans. It makes a huge difference if you are doing a computer vision algorithm that you have 5 man years of software development and the sensor’s sensitivity changes by even 1%.
Furthermore, the form factor of webcams change frequently as well. This doesn’t make a difference when it is just on your desk. It makes a huge difference when your camera and lens is fixtured in a machine that has 500 hours of CAD work to design, much less build. Moving the camera and lens 10cm might not be possible!
Do you need to choose your lens?
Webcams come with an integrated lens that is suitable for general viewing, and this lens is integrated with the camera and not changeable. Industrial machine vision cameras come with no lenses as not only do lenses come in a variety of focal lengths for different magnification, but also lenses coming in a variety of resolutions. Choosing a lens requires you to know the size of the sensor, your working distance, your field of view, and the pixel size. (See related educational blogs on lenses at end of this post)
What are your options for a low cost camera solution?
If you need industrial machine vision camera solutions with a solid SDK, long life cycles, at a low price, there several solutions to consider. Rolling shutter imagers are always lower price which are always a place to start along with USB2 interfaces. Read our previous blog HERE which outlines some specific models which are low cost. There is also a great new platform coming providing 5 Megapixel resolution with a rolling shutter imager, but with great performance for $280! Contact us for more details.
These days, there are probably over 30 different camera manufacturers making products for the machine vision marketplace, many using the identical image sensors. Considering anyone can just put up a web page and start selling, how can a user know which product to purchase? Are there really any differences? Should I just purchase on price? Should I buy from a distributor or from the manufacturer direct? All great questions, that we will attempt to answer.
The very simple and quick answer is that if you just need to get a image in good light, pretty much any camera from any vendor will do that job. No matter if it is from a large company or a 2 person startup, when you take the product out of the box, you should see a “good” image.
But if you said, “I want this camera to run 24/7 for the next 5 years, I want to be able to develop complex software to integrate into my machine, I need the image sensor plane to be within a certain tolerance for each machine,” this changes the situation.
The reason we have chosen the camera manufacturers we sell products from is because each of them has a proven track record of reliability. Each of the companies we represent ships 6 figures of cameras per year. IDS Imagingfor instance ships close to 200,000 cameras per year and has a return rate of under 0.3%.
Allied Visionwas the first company to incorporate the Precision Timing Protocol (PTP) which allows for precise multi camera sync, enabling our clients to not only make sure the application will work, but it doesn’t take years to develop it.
Teledyne Dalsa, besides being a leader in line scan technology, has a SDK that has been built upon for over 30 years.
JAI’s prism technology is so good that its competitors actually have JAI
manufacture for them. This isn’t to say there aren’t other camera companies with such characteristics. There are, and many of the other camera companies have excellent products as well. It is just that we have chosen these companies, and we have stood with them for 15+ years for good reasons… sleepabilty!
What does this mean to you as a client? Yes, you can purchase a camera from any vendor, which on a $500 camera, you might even be saving $50 a camera. If you purchase 100 cameras a year, this adds up to a reasonable savings of $5,000. But what happens if your machine, which you sell globally, has a camera that fails. What is the cost of the line going down at your client? How do you look in your client’s eyes? What is the cost for you to fix it?
If you had your choice of buying a camera with the same characteristics at roughly the same price, but one company makes 20,000 a year, and the other makes 200,000 a year, which would you choose? The same is true if you can choose between a company that has offices all over the world, or just in one country. Or one that has many application engineers to answer your questions, or just one.
We just want to point out that if you are making a purchase on price alone, depending upon your circumstances, it might not really be a savings at all. In fact, it might actually not only be costing you money, but it might even be costing you your sleep!
The AR1820HS Image sensor 18 mega pixel sensor in the IDS Imaging IDS Imaging UI-3590 camera models was launched by the sensor manufacturer ON Semiconductor as a pure color sensor. Like all color sensors, the Bayer filter means that you get color images with effectively only around a quarter of the nominal sensor resolution, as the color information for each pixel is obtained from four neighbors
To use each individual pixel, however, it is not sufficient to operate the sensor in RAW data format (without Bayer interpolation). Operating in a raw format results in a different brightness perception of the individual pixels and NOT a usable image.
This technical tip will show you how to use the color sensor as a “pure” mono sensor by appropriate parameter settings and the use of suitable light sources, in order to obtain a significantly higher resolution.
Applications which require a color image and precision will benefit from this camera and method.. and all for less than $600 with the IDS UI-3590LE camera
The principle of digital image sensors such as the 18 MP ON Semiconductor AR1820HS means that they acquire only brightness, but not color information.
As a result, a color filter is applied to each pixel during manufacture of the color sensors. This is known as the Bayer matrix.
Of each four pixels, two pixels are given a green color filter, one pixel a red filter and one pixel a blue color filter. This color distribution corresponds to the perception of the human eye and is referred to as the Bayer matrix.
A pixel depicts only the information for one color.
To obtain the complete RGB values for each pixel, the missing primary colors are interpolated from four neighboring pixels using appropriate algorithms. This color interpolation assumes that there are only slight color differences between two adjacent pixels of the same color. Strictly speaking, a sensor with Bayer matrix therefore has only a quarter of the native sensor resolution
Sensor mono mode
Although the Bayer matrix cannot simply be rendered invisible for mono mode, the following two solutions show how you can achieve the desired result depending on the type of application.
1) For “grey scenes” (i.e. dark pin on a white background)
If the 18 MP color sensor is to be used in mono mode for achromatic scenes, note that a broad band light source (white light) must still be used. This is because of the sensor’s Bayer matrix. With this sensor, monochromatic (single color) light would have resulted in the individual pixels transmitting no or less information depending on the wavelength of the light, due to the RGB filter layers used (see Figure 2). This can result in a different brightness perception for the individual pixels. In this case, the RGB enhancement must be calibrated separately for R, G, and B. As a result, you then obtain an identical brightness perception for all pixels as with a mono sensor.
Note: This RGB calibration is only valid for this specific light source and a “grey” scene. If the light source (wavelength) changes, the RGB enhancement factors have to be re-adjusted.
Contact 1stVision to obtain instructions on switching the Bayer matrix to “invisible” using the uEye Cockpit:2) For “color or grey scenes”
If you work with color scenes in your application, the brightness sensitivity of the individual Bayer pixels constantly changes with the variation in color components. There is also a way to achieve genuine mono mode in this situation. The solution lies in the color spectrum of the 18 MP ON Semiconductor AR1820HS.
Above a wavelength of around 900 nm the color filters for the individual pixels have similar spectral properties. Beyond this threshold, all pixels on the sensor respond practically identically to incident light again – exactly as with a dedicated mono sensor. This means that the Bayer matrix can also be made invisible using this method, both for color and also for grey scenes
In order to be able to use this spectral property of the sensor as described, you must observe the following:
Ensure defined lighting conditions, i.e. seal off light with other wavelengths shorter than 900 nm as far as possible.
Order your uEye camera with AR1820HS sensor specifically with GL filter (glass). The HQ filter normally ordered with this sensor would block out the long-wave light. By contrast, the GL filter allows light beyond 900 nm to pass with high transmission. The highest possible signal strength thus arrives at the sensor.
Full Information on the IDS Imaging UI-3590LE and UI-3590CP cameras can be found below