The rover "Curiosity": an explanation of the choice of 2 megapixel cameras
One of the most popular topics of the past week is undoubtedly the successful landing of the “Curiosity” apparatus on the surface of Mars. In this article, which is a translation of an interview with one of the leaders of the Mars Science Laboratory mission , I would like to reveal a few details about the photographic equipment, which is located on the rover.
Several factors influenced the choice of 2 megapixel sensors for use in cameras of the NASA “Curiosity” rover , as the project manager for their development claims. “The low data rate available for sending images back to Earth, as well as knowledge of the sensor family among the development team, have played a role,” says Mike Ravin, who works at Malin Space Science Systems . “But the main factor is the specification, which was approved back in 2004.” The cameras are capable of producing high-resolution panoramas, but not three-dimensional films that director James Cameron wanted to receive.
“There is a popular belief that projects like this should be very progressive, but there are things that play against it. These ideas were expressed in 2004, and you cannot offer one specification, and then abandon it and start developing something else. A 2 megapixel camera with 8 GB of memory was not bad for 2004. Of course, today it doesn’t compare with what the iPhone has. ”
A Mastcam camera with a 34mm lens (equivalent focal length - 115 mm) from the Curiosity rover, developed by Mike Ravin’s team at Malin Space Science Systems.
')
The first consideration we had in mind when choosing a sensor was the volume of data produced by it.
Curiosity has three ways to transfer data back to Earth, but only a transmitter at a frequency of decimeter waves can be used to send the amount of data that is contained in the images. “The decimeter antenna transmits a signal to two spacecraft orbiting around Mars, which send it further to Earth. This way we get the most data. We receive about 250 Mbit of information per day, and this volume should include information from various devices, so that a small range is allocated for cameras. ”Explains Ravin.
Another factor was that the same type of sensor had to be used in four different cameras (MAHLI, two Mastcam and MARDI). “Anything in such projects has its value and carries real and intangible risk. Cameras are distinguished by optics, however, creating them on the basis of a single platform, we did not have to test and prepare them separately. This made it possible to make the solution more reliable and less expensive than if it had to be developed four times. ”
“We developed all four cameras around a common architecture, and when choosing a sensor we made a start from this. We wanted to be able to take pictures with high frequency, in particular with the camera used during landing. ”Explains Ravin. - “MARDI, downward camera, worked only two minutes while descending to the planet, so the speed of the shooting was so important. KAI-2020 was Kodak's smallest chip capable of shooting 720p HD video. We also considered the 4 megapixel version, but it would have worked twice as slowly. In addition, CMOS sensors were not yet credible in 2004. It is now they are an interesting option, but not then. ”
The preview image from the MARDI landing chamber captured the falling heat shield at the initial moment of the final landing phase.
Another advantage of the Truesense image sensors was that the team was familiar with their characteristics. “We had experience gained from decades of working with Kodak sensors (now Truesense). We knew well how to work with the clock frequency of the sensors: these CCD sensors are very easy to set up, ”says Ravin. “We also had to be sure of the memory of the cameras: in the end, we stopped at that flash memory, because she passed many tests for radiation. ”
The image with the maximum resolution, published on August 8, gives a clearer picture of the capabilities of the 2 megapixel cameras installed on board the CurieCity.
“And a small number of pixels by today's standards should not be a problem, since the Mastcam system is capable of gluing frames. Such an installation, performed by a camera with a large number of pixels, would not give us great advantages compared to what we have now. ”
Ravin is clearly proud of the result of his team’s work during this project, and notes: “We are already receiving images from Mars.” But his voice clearly shows frustration when he mentions zoom lenses that were prepared for the project: “Now they lie in a container at the other end of the building. ”
Sadly, the refusal to use these lenses in the project means that the material with three-dimensional scenes, which the team hoped for, will not be filmed. “These were supposed to be lenses with a focal length in the range of 6.5-100 mm, which, after adjusting them to one distance, would allow us to get stereo images,” says Ravin. However, problems arising from the inability to use liquid lubricant (which would have required the use of a battery to ensure uninterrupted operation at low temperatures on Mars) proved to be intractable, and development was suspended.
Advantages of image splicing: a two-frame panorama based on 1024 by 1024 pixel pictures taken by the navigation cameras of the rover. The Mastcam system should also produce higher resolution color frames.
Director James Cameron was a member of the team and with his conversations about three-dimensional films in which “Curiosity” moves along the Martian landscapes, was able to convince NASA to resume this project. Sadly, although the project was launched again, even Cameron’s enthusiasm was not enough to complete the development on time. “We have to get great editing pictures even with the lenses that we have, but it’s not possible to get widescreen cinematographic scenes or 3D films from Mars that Cameron wanted,”
Another view is the image of the rover and its environment in the polar projection. The photo is assembled from low-resolution frames, with the exception of the two above.
Several factors influenced the choice of 2 megapixel sensors for use in cameras of the NASA “Curiosity” rover , as the project manager for their development claims. “The low data rate available for sending images back to Earth, as well as knowledge of the sensor family among the development team, have played a role,” says Mike Ravin, who works at Malin Space Science Systems . “But the main factor is the specification, which was approved back in 2004.” The cameras are capable of producing high-resolution panoramas, but not three-dimensional films that director James Cameron wanted to receive.
“There is a popular belief that projects like this should be very progressive, but there are things that play against it. These ideas were expressed in 2004, and you cannot offer one specification, and then abandon it and start developing something else. A 2 megapixel camera with 8 GB of memory was not bad for 2004. Of course, today it doesn’t compare with what the iPhone has. ”
A Mastcam camera with a 34mm lens (equivalent focal length - 115 mm) from the Curiosity rover, developed by Mike Ravin’s team at Malin Space Science Systems.
')
The first consideration we had in mind when choosing a sensor was the volume of data produced by it.
Curiosity has three ways to transfer data back to Earth, but only a transmitter at a frequency of decimeter waves can be used to send the amount of data that is contained in the images. “The decimeter antenna transmits a signal to two spacecraft orbiting around Mars, which send it further to Earth. This way we get the most data. We receive about 250 Mbit of information per day, and this volume should include information from various devices, so that a small range is allocated for cameras. ”Explains Ravin.
Another factor was that the same type of sensor had to be used in four different cameras (MAHLI, two Mastcam and MARDI). “Anything in such projects has its value and carries real and intangible risk. Cameras are distinguished by optics, however, creating them on the basis of a single platform, we did not have to test and prepare them separately. This made it possible to make the solution more reliable and less expensive than if it had to be developed four times. ”
“We developed all four cameras around a common architecture, and when choosing a sensor we made a start from this. We wanted to be able to take pictures with high frequency, in particular with the camera used during landing. ”Explains Ravin. - “MARDI, downward camera, worked only two minutes while descending to the planet, so the speed of the shooting was so important. KAI-2020 was Kodak's smallest chip capable of shooting 720p HD video. We also considered the 4 megapixel version, but it would have worked twice as slowly. In addition, CMOS sensors were not yet credible in 2004. It is now they are an interesting option, but not then. ”
The preview image from the MARDI landing chamber captured the falling heat shield at the initial moment of the final landing phase.
Another advantage of the Truesense image sensors was that the team was familiar with their characteristics. “We had experience gained from decades of working with Kodak sensors (now Truesense). We knew well how to work with the clock frequency of the sensors: these CCD sensors are very easy to set up, ”says Ravin. “We also had to be sure of the memory of the cameras: in the end, we stopped at that flash memory, because she passed many tests for radiation. ”
The image with the maximum resolution, published on August 8, gives a clearer picture of the capabilities of the 2 megapixel cameras installed on board the CurieCity.
“And a small number of pixels by today's standards should not be a problem, since the Mastcam system is capable of gluing frames. Such an installation, performed by a camera with a large number of pixels, would not give us great advantages compared to what we have now. ”
Ravin is clearly proud of the result of his team’s work during this project, and notes: “We are already receiving images from Mars.” But his voice clearly shows frustration when he mentions zoom lenses that were prepared for the project: “Now they lie in a container at the other end of the building. ”
Sadly, the refusal to use these lenses in the project means that the material with three-dimensional scenes, which the team hoped for, will not be filmed. “These were supposed to be lenses with a focal length in the range of 6.5-100 mm, which, after adjusting them to one distance, would allow us to get stereo images,” says Ravin. However, problems arising from the inability to use liquid lubricant (which would have required the use of a battery to ensure uninterrupted operation at low temperatures on Mars) proved to be intractable, and development was suspended.
Advantages of image splicing: a two-frame panorama based on 1024 by 1024 pixel pictures taken by the navigation cameras of the rover. The Mastcam system should also produce higher resolution color frames.
Director James Cameron was a member of the team and with his conversations about three-dimensional films in which “Curiosity” moves along the Martian landscapes, was able to convince NASA to resume this project. Sadly, although the project was launched again, even Cameron’s enthusiasm was not enough to complete the development on time. “We have to get great editing pictures even with the lenses that we have, but it’s not possible to get widescreen cinematographic scenes or 3D films from Mars that Cameron wanted,”
Another view is the image of the rover and its environment in the polar projection. The photo is assembled from low-resolution frames, with the exception of the two above.
Source: https://habr.com/ru/post/149565/
All Articles