High Speed Camera

Table of Contents 

An image taken with a high speed camera that captures a bullet going through an apple [1]. The point of impact can be well observed.

Image of bullet going through an apple being captured by a high speed camera. The point of impact and the bullet in mid-air can be captured only at high speeds.

Introduction to Cameras

A camera is a light-tight box that allows light to enter through a small opening, also known as the aperture and hits a sensor. This device uses mirrors, optics, and special sensors to be able to capture an image and display it as computer pixels. Pixels contain the information for a photo - the RGB values (0-255, 3 layers) [2]. 

There are different types of cameras: phone cameras, digital SLR cameras, mirrorless cameras, and point and shoot cameras. 


Basic Photography Terminology


Video explaining the basics of photography in less than ten minutes [3].


Below are some common terms that are important to control when using a camera and the benefit with a high speed camera [4]:

  1. Shutter speed: The speed in which an image is being captured. 
  2. Aperture: This is the opening in a lens that allows light to pass through when opened. It is numbered from the lowest aperture (around F3.5 usually) to the highest (F36). A low aperture can be used to achieve this speed even in low light conditions.
  3. Exposure: The amount of light that reaches the camera sensor. Overexposure occurs when there is too much light in the frame and the photo will appear too bright if not fully white. Underexposure occurs when the opposite occurs, not enough light and will be dark or black. Being able to control the camera settings (e.g. aperture, shutter speed) to find the right balance is key, especially for object detection and identification.
  4. ISO: This refers to the sensitivity of the camera. Having a high ISO value (e.g. ISO 6400) will mean the camera is highly sensitive. Most importantly, the high speed camera will be able to capture photos in low light environments.
  5. Zoom: This offers users to look at objects closer or farther away from the camera. Most cameras have a zoom lens offering these optical zoom capabilities. On the contrary, some cameras may not and will have "digital zoom" which means it will crop the photo rather than using the lens to zoom in, resulting in lower quality image (a cropped image).

In industry, having a fast shutter will allow for many frames to be captured and give detail of what is happening. 


Sensor Size and Aspect Ratio

The image sensor is what is used to capture the light coming into the camera. They use photo sensors to turn the light into digital information. Due to this, the sensor in a camera can play a major role in determining how the captured image looks. Sensors within cameras vary based on their sizes and aspect ratios. Typically, the larger the sensor, the more light it can capture, and thus, it produces a sharper and clearer image. The industry standard are full frame sensors that are 36mm in width and 24mm in height. Cameras with sensors smaller than the industry standard are called cropped sensors.

The aspect ratio of the camera is also associated with the size of the sensor. This is because the aspect ratio is the ratio between the width and height of the sensor within the camera. Images that are captured on cameras with a different aspect ratio than the device meant to display the image, may need to be cropped or resized [5].


Frame Rate vs Shutter Speed

Frame rate is defined as the speed of a frame in a given time interval [6]. This is quantified in terms of frames per seconds or FPS. The most common frame rate for cameras are 24 and 30 FPS [6]. Other frame rates that are present for more expensive cameras are 60 FPS and 120 FPS. High frame rate cameras are needed in applications that have objects moving at high speeds in order to collect more accurate data. 

On the contrary, shutter speed is the time the aperture is open to allow light to reach the sensor. Frame rate and shutter speed are related to each other, but different in what they do. 

For the golf project, obtaining a fast shutter speed was important to capture the golf ball. Having a clear image compared to a blurred photo from movement. A high frame rate was used (60 FPS), and the video was slowed down in the software program to allow for the golf ball image to be seen in successive frames. It is important to have a high frame rate when selecting a high speed camera.


What is a High Speed Camera?

A high speed camera's main characteristic is a fast shutter speed. High speed cameras usually have a shutter faster than a thousandth of a second. This is able to capture short duration events, such as a golf ball moving mid-air, or a bullet going through an apple as pictured above. The capture is clear and does not have significant motion blur.

An example of a high speed camera is the Sony DSC-RX100M7, which was used for the golf project. The fastest shutter speed for this camera is 1/32,000 of a second [7]. As well, it has a wide aperture of F2.8 allowing for more light to enter. Other brands and models may offer different speeds and technical features, such as different aperture size and fast shutter speed. For more on the Sony camera and its use for the golf project, Golf Project Cameras.

Slow motion ("slo-mo") requires the frames to be played at a slower speed. Usually 2.5 to 5 times slower than the frames per second set. This is either done on the camera using a special setting or done after using a video editing program.


Image Processing

Image processing is the technique of extracting, analyzing, or enhancing the data provided by an image [8]. Image processing has a great deal of use cases, some of which are outlined below:

Dataset Creation

Similar images can be compiled to create datasets in varying formats. These datasets are usually based around particular objects or items people may be interested in identifying within the image. Examples of this are datasets based on images containing dogs or cars. Some of the popular format standards for datasets are Pascal VOC and COCO formats. There are also open source tools such as CVAT and LabelImg that can be used to create custom datasets with relative ease. These datasets are then used to train Deep Neural Networks (DNN) so that they can detect whether the objects they are based on are present with another image or video feed.

Data Analysis

The raw data from images can also be very useful in a lot of cases. The main information within an image is the array of data that contain its red green and blue (RGB) values. Therefore, analyzing that data using software has many applications. An example of this is using MATLAB to compare the RGB values of two different images. This can be used to assess how alike the two images are.

General Photo Editing

This can be done using software tools such as Adobe Photoshop and Lightroom. General photo editing can involve white balance changes, contrast changes, cropping and much more. It may also be necessary to edit photos to generate vector images so that they can be used to generate tool paths for applications such as laser cutting or CNC machining.


Exporting and File Types

The most common file types for a camera are JPEG and RAW for photos and .MOV or .mp4 for movies. RAW files are unprocessed and contain information on the photo. They are usually larger in size than a JPEG image and require specialized photo editing software, such as Adobe Photoshop. These settings will need to be specified on the camera menu.

When exporting use a camera to USB 2.0 adapter or use the SD card if you have an SD card reader. Photos can be transferred to computer for processing and analyses, as well as better viewing. Some SD cards will offer a bluetooth connection which is an alternative option for transferring files.


References

[1] S. Dunne, “'Science In Motion' Photography At Bruce Museum,” courant.com, 22-Aug-2016. [Online]. Available: https://www.courant.com/entertainment/arts-theater/hc-bruce-scientific-photography-0825-20160822-story.html?trb=20181212. [Accessed: 05-Feb-2021].

[2] “The Computer Vision Pipeline, Part 2: input images,” Manning, 18-Aug-2019. [Online]. Available: https://freecontent.manning.com/the-computer-vision-pipeline-part-2-input-images/. [Accessed: 11-Feb-2021].

[3] P. McKinnon, “CAMERA BASICS!” [Online]. Available: https://www.youtube.com/watch?v=ixRKeQMa7Nc&feature=emb_logo. [Accessed: 11-Feb-2021].

[4] P. T. Editor, “Basic Camera Settings for Real Estate Photography Exteriors,” PhotographyTalk, 25-Jul-2020. [Online]. Available: https://www.photographytalk.com/real-estate-photography/9060-basic-camera-settings-for-real-estate-photography-exteriors. [Accessed: 11-Feb-2021].

[5] D.P. Curtin, “Image Sensors-Sizes and Aspect Ratios,” ShortCourses. https://www.shortcourses.com/guide/guide2-9.html#:~:text=Image%20sensors%20come%20in%20a%20variety%20of%20aspect%20ratios%E2%80%94%20the,fall%20in%20between%20these%20extremes [Accessed: Mar. 15, 2021].

[6] Sam Morrill Posted on July 17, About The Author Sam Morrill, A. T. Author, and S. Morrill, “Frame rate vs. shutter speed, setting the record straight,” Vimeo Blog, 05-Jun-2020. [Online]. Available: https://vimeo.com/blog/post/frame-rate-vs-shutter-speed-setting-the-record-str/. [Accessed: 08-Feb-2021].

[7] “Sony RX100 VII Compact Camera, Unrivalled AF,” Sony. [Online]. Available: https://www.sony.ca/en/electronics/cyber-shot-compact-cameras/dsc-rx100m7. [Accessed: 11-Feb-2021].

[8] The Editors of Encyclopedia, “Image processing,” Britannica. https://www.britannica.com/technology/image-processing [Accessed: Mar. 16, 2021].

Contributors:

UserLast Update
Mayurakhi Khan 931 days ago
d327pate@uwaterloo.ca
Former user (Deleted)
Former user (Deleted)
Former user (Deleted)

Faculty Advisor: Chris Rennick