The use of slit-scan photography is actually quite old. It is often called line-scan, photo finish, or streak photography. Slit-scan photography has a rich and colorful history rooted in chemical analog photography. This technique is often used to visualize high-speed events such as missiles and bullets, although it is probably best known as photo finish photography used to determine the outcome of races.
In the past, slit-scan photographic systems used a sheet of film that was moved past a slit. These cameras were most commonly used as photo finish cameras at races and could very precisely measure the time one horse might have won the race by, for example. There were a number of designs of these systems. One of the most interesting slit-scan cameras had the camera and film moving at the same time to create a panoramic picture. The last camera on the market to use this technique was the Spinner Dolphin 360 made by Lomography.
Modern times have replaced film with digital sensors and slit-scan photography is more popular than ever, although most people have never seen the results. Slit-scan imaging solves a number of problems found in industrial applications. Quality control inspections found in high-speed assembly lines can look for product defects in real time. Industrial applications are the most common uses of the technology today. Although it not the most popular form of photography, the resulting images are often a surprise to the photographer and can be quite stunning and beautiful in their own right.
To make a slit-scan image a large number of images are collected in video format. To create a slit scan image, one row of pixels is extracted from each frame of the video and placed adjacent to the same row from the next frame. Each row of pixels represents a duration of time. If the video was recorded at a rate of 30 frames a second, then the row of pixels will represent 1/30th of a second. The resulting image is really a representation of motion and time. To figure out how much video to collect, count each frame of the video as one pixel in the width of the image to be part of the final image. In this case, I collected 4 x 60 x 30 or a grand total of 7200 images. Thus my final image was 1920 high and 7200 pixels wide. The 1920 comes from the video frame placed vertically to get a larger pixel count in that direction.
To extract the slit-scan image out of the finished MOV file I use a useful program written by Martin Dixon called Slit-Scan. It is a free download and is available from this website. The program is available for both Mac and Windows operating systems and requires the programming environment called Processing to running. This is by far the easiest way to make slit-scan images currently.
To measure the results of a toy car race, students in my class recorded two Matchbox cars going down a track. The video was recorded at 1000FPS with an Edgertronic camera.
A typical photo finish image of a Hot Wheels car race. The red car crossed the finish 114 pixels in front of the blue car since each pixel in the horizontal direction represents 1/1000th of a second, the red car won the race by .114 seconds. The slit-scan image is a very powerful tool and useful for making image measurements. If the length of the car is known, then the pixels can be measured in the horizontal direction to determine the amount of time for the car to pass a fixed point. From this information, the velocity of the car can be calculated.
Ocean waves are recorded when moving. Since each pixel in the horizontal direction represents 1/30th of a second, the wave motion can be measured.
If the camera is rotated, the extracted slit scan becomes a panoramic image. Here a junction in the college’s hallways is imaged by Nate Dileas, Scott Semler, and Makayla Roof from my high-speed imaging class. Since the camera was moving very slowly, Nate ducked around the camera and was recorded twice. This is a classic strategy for this technology and dates way back. If you look at old panoramic images you will almost always see at least one individual that is included twice in the photograph.
In this picture, a colleague Dan Hughes volunteered to sit patiently on a rotating chair. The resulting image is a peripheral slit scan image sometimes called a peripheral portrait and reveals the full circumference of his head. This technique has been used to record Roman vases, and tread wear patterns on tires.
Instead of a human subject, a peripheral slit scan image is collected of a rotating Dahlia flower. The flower has the slit of pixels extracted from the video is parallel to the axis of rotation. If the extracted row of pixels is not parallel or taken at an angle the resulting images can be quite weird and surprising.
The same dahlia flower as above is used to make a slit scan image, but here the row of extracted pixels is at a 90-degree angle to the axis of rotation. I personally find these images new and exciting to make. Even after making these images for 20 years, I still never quite know what to expect.
An off-axis image of another Dahlia flower this image was extracted from a full resolution .mov file taken with a Canon 5DMkIII.
The same dahlia flower movie file used to make a peripheral slit scan image of the flower.
A bouquet of spring tulips imaged off-axis make a unique twisty picture. The flowers were placed on a turntable that took three minutes to make one complete revolution. This image was collected in camera by using a Better Light scan back camera which had the scanner parked in a fixed position. The camera has a feature that allows the taking very high-resolution panoramic images. The full image requires approximately five minutes to collect because the camera’s operation itself is slow.
Here a dahlia flower is imaged with the axis of the rotation and the camera offset by 15 degrees. Strange effects are quite common with this technique.
Downtown Richmond VA is imaged by collecting a high-speed video from an iPhone 6 pointed out the window of a moving car. The handheld image shows the aspect ratio is not correct – the car should have been moving slower. Slit-scan imaging is commonly used to map the ground from high flying aircraft. Surprisingly, aerial slit scan still uses large and long rolls of film even in 2017.
Slit-scan is used to record the patterns on a cone seashell. Another example of peripheral streak imaging.
The technique is used to capture the arrangement of corn kernels in Flint corn. Flint corn is also known as Indian corn or calico corn and is a common decorative corn seen here in the United States throughout the U.S.’s Thanksgiving holiday. This image shows the variation of the placement and color of the kernels of corn around the full corncob in one image.
Many images just look strange like this slit scan image of a historic flashbulb firing. The slit image is used to determine the timing of the flash here the M3 flash bulb is brightest for about 20 milliseconds. As in all of the slit images time goes from left to right here.
An abalone paua shell from New Zealand is imaged as an off-axis slit scan image. The constant pattern at the bottom is due to the rotation stage being turned off.
Have you ever wondered what beach glass placed on a rotation stage placed on a second rotation stage imaged in off-axis silt scan photography would look like? I have and pictured above is the result. Oddly enough, this image will be familiar to readers that are used to looking at rotation charts for the moons of Jupiter which are displayed in a similar pattern.
For a number of years, I have been intrigued by the unique patterns displayed by objects in a variety of ways through the use of slit can imagery. I hope the interested reader will give this technique a try.
About the author: Ted Kinsman is an assistant professor of photographic technology at the Rochester Institute of Technology. He teaches advanced photographic technology, light microscopy, and macro photography courses. Kinsman specializes in applying physics to photography. You can find more about him and his work in his faculty profile and on his website.
from PetaPixel http://ift.tt/2xP4A9b
from WordPress http://ift.tt/2xOU1TL
No comments:
Post a Comment