How HiRISE Works
Lesson Two: Resolution and Binning
What is Resolution?
HiRISE is the highest spatial resolution camera ever flown to Mars. But what is spatial resolution? We use this phrase to mean the ability to distinguish ("resolve") different objects based on their size and shape ("spatial" characteristics).
Spatial resolution is the minimum distance between two objects that are distinguishable. This is always a good thing to think about when you're trying to image a certain type of object (boulders vs. volcanoes vs. craters etc.). Although this can sometimes be a subjective measurement, it is closely related to something called the "Ground Sampling Dimension" (GSD). The GSD is the length that a single pixel would take up on the ground and for HiRISE is measured in units of cm/pixel. For instance, when HiRISE is operating at its highest resolution, its GSD is approximately 30 cm per pixel. However, several pixels (usually 3 or 4 together) are required to resolve features on the surface of Mars. Therefore, HiRISE is able to resolve features that are about 1 meter across, hence one-meter spatial resolution.
Other instruments onboard the MRO spacecraft specialize in other types of resolution. For example CRISM, with less spatial resolution than HiRISE, uses infrared colors to help distinguish objects and has very good spectral resolution. HiRISE has some spectral resolution because we can take pictures in three colors. A third kind of resolution, called temporal resolution comes when one distinguishes objects based on changes in time. The MCS and MARCI instruments onboard MRO use high temporal resolution to track changes in the Martian weather. As you can see, the MRO mission has instruments with new capabilities for every kind of resolution!
Getting the Best Picture
There are a lot of different factors that affect the quality of a picture. If you've ever shopped for a digital camera you have heard that more pixels are better. Indeed, the more pixels you have over the same picture, the more detail you can see. The rule of thumb is that you need 3 pixels across an object to be sure that it is really there and roughly 10 pixels across it to know what it is. We know this too, so HiRISE doesn't skimp on pixels! Today, you can typically buy a 4-8 "Megapixel" digital camera (which gives you a very respectable 4-8 million pixels). HiRISE, on the other hand, can take 1,200 Megapixel black and white pictures and 240 Megapixel color pictures.
But a large number of pixels alone isn't enough to give you a good picture. You also need the picture to be in focus. As you would expect, HiRISE has optics that have been polished to extreme precision. As discovered by Isaac Newton in the 1600's, you get less distortion of the colors if you use mirrors instead of lenses. Therefore, you will find 5 mirrors but no lenses in HiRISE. We don't like to take any chances, so we also have a mechanical focus mechanism. Our test pictures of stars show that the optics are working fabulously and are focusing the light with exquisite precision.
But that's not all! If you want to take pictures of small things from far away, you are going to need a lot of 'zoom' or magnification. In the case of HiRISE, we are orbiting about 300 km (or 190 miles) from the surface of Mars and we want to see objects only a meter (or yard) across. As we mentioned earlier, this means we want at least 3 pixels per meter (or yard). Therefore, each pixel should cover no more than about 30 cm (or 1 foot). In order to do this, we needed a "zoom lens" (really the camera and mirror system) the size of a desk! This is why HiRISE is the largest camera ever to fly to Mars.
Wait there's more! If you shake the camera while you're taking a picture, it will come out blurry. Professional photographers use tripods to hold their cameras. But we have to take a picture of Mars as it whips by at almost 4 kilometers (2 miles) per second! HiRISE has to be moved with extreme precision and smoothness to track the surface of Mars. Luckily for us, HiRISE is in the able hands of the MRO spacecraft, which was built to accomplish this daunting task. Still, we worry that even little motions, such as moving the solar arrays, will make the spacecraft jiggle and blur our pictures. Therefore, when we want to take our sharpest pictures of Mars, we have to (politely) ask all the other instruments and the spacecraft to be extra quiet... not something we can demand all the time!
Finally, there is the problem of getting the right exposure. You may have noticed that if you take a picture that comes out too dark, the picture is 'grainy'? Or if its too bright, the picture is 'washed out'? You just can't see as many details if you set the exposure wrong. We actually have two ways to adjust how much light we gather into each pixel of the final picture. One is called "Time Delay Integration" (TDI). TDI is a bit tricky to explain, but the short version is that we can chose to gather 8, 32, 64, or 128 times as much light as a single pixel would. The other method is called binning and is explained below.
How Does Binning Work?
With binning, instead of getting information from each pixel in the camera, groups of pixels are combined to form larger single pixels, as seen in Figure 1. For instance, when there is 2x2 binning, you receive information for a single pixel where before the binning you had 4 pixels (a box with sides 2 pixels long). Each new pixel you receive contains all the light from the 4 original pixels. This makes that pixel 4 times brighter, but you also have 4 times fewer pixels. This is all done within the camera's electronics, based on how we command HiRISE at the time the picture is taken. We do not change the binning afterward. Click here for an example of 4x4 binning on a Mars Observer camera image.
Since binning reduces the spatial resolution, why would be ever want to use it? If the scene is too dark, our experience shows that the picture will be so grainy that you won't be able to see the finest details anyway. So you lose nothing by binning dark images to get the exposure right. There are other cases where binning is useful. If the weather is hazy, or if the spacecraft is jittering, then the picture will be blurry anyway. Again, you lose nothing by binning. Remember that we can't ask the spacecraft to be quiet all the time, so we expect to have to bin many of our images. Finally, we just can't deal with too many giant 1,200 Megapixel pictures! In many cases, it's better to take a smaller picture with fewer pixels. We can do this by taking a picture that covers a smaller part of Mars, or we can use binning. In some cases, we simply don't need to identify all the things as small as 1 meter (1 yard). Then we can use binning to get a picture with just the right amount of spatial resolution and not clog up the system with pixels we won't need. Limiting the amount of data is important since the camera will be restricted by the rate at which data can be radioed back to Earth to returning approximately two full resolution images a day. Binning allows more images to be returned.
How Does Binning Affect My Image Suggestion?
When you suggest an image observation for HiRISE on HiWeb, you will be asked to choose a degree of binning. Your first inclination may be to specify no binning in order to achieve the highest possible spatial resolution. However, unless you are interested in analyzing the finest details of your area, for example, looking at very small geologic features on a bright surface on a very clear day you may not actually need all those pixels. In addition, all those pixels will jam up the data link back to Earth and prevent other pixels in other images from being sent home. For that reason, the HiRISE team may give your suggestion a lower priority. Therefore, unless you are interested in the smallest details of the features in your suggested image and can justify your request for full-resolution, you should probably select at least 2x2 binning.
How Can I Decide What Degree of Binning is Best?
One way to start thinking about the resolution you need for an image is to find out if there already is a narrow angle (high-resolution) Mars Orbiter Camera (MOC) image of your target. HiRISE's highest resolution will be five times better than MOC's best resolution and will also provide color information. If there is not a MOC image of the target, it may be better to suggest moderate binning to produce a first look at the surface. Alternatively, you could suggest MOC itself targets your site through their Mars Orbiter Camera Target Request Site. As of January 2006, MOC is still operating aboard the Mars Global Surveyor spacecraft orbiting about Mars.
If there is an existing high-resolution MOC image of the area, try to get a general idea of how much better resolution you would need in a HiRISE image to understand the surface. Likewise, you may also want to look at the THEMIS daytime visual images from the Mars Odyssey Spacecraft. These are available at 18m/pixel and help give a better idea of the context of the region. Both the THEMIS and MOC images are integrated into the HiRISE image suggestion facility viewer for you to peruse. These data are available by simply clicking on the radio button on the HiRISE image suggestion facility.
To learn about the basic workings of HiRISE, go back to Lesson One.
Back to HiRISE Learning Page.