Competition have transformed drones from an expensive hobby into an efficient commercial technology and great entertainment.
The fall in prices has caused a huge demand for drones. Of course, companies responded adequately by actively expanding a new direction on this platform. As a result, modern drones are already actively used for area reconnaissance, search, guard, delivery, in drone racing, etc.
But, of course, FPV (first-person view) technology has become one of the main directions of development. The growth of its popularity began in the early 2010s. This technology enables wireless transmission of data from the onboard camera via VTX to FPV goggles or video monitor in real time. FPV very closely mimics the feeling a real pilot. FPV very accurately simulates the sensations of a real pilot, providing vivid emotions and allowing people to realize their dreams of flying in the sky.
Of course, the camera is one of the main components of this system. Its optimal choice will ensure maximum enjoyment of virtual flights at minimum cost. Therefore, information about their features can simplify this task.
Of course, the image sensor is the main component of camcorder. It converts the image from the lens sensor into an electrical signal. All modern cameras use CCD (charge-coupled device) or CMOS (complementary metal-oxide-semiconductor) sensors.
The CCD is an analog sensor. When light is absorbed, each pixel of the matrix accumulates charge and converts it into a video signal voltage, which is proportional to its illumination. The minimum number of intermediate transitions and the absence of active elements provide high accuracy of CCD sensing elements.
The CMOS sensor is a digital device with Active Pixel Sensor. Each pixel uses its own amplifier, which converts the charge of the sensing element into voltage. This technology provides control over every pixel.
CCD technology was invented at Bell Labs in 1969. Since then, its size has been decreasing while the number of sensing elements has increased. Of course, the size of each pixel also decreased. As a result, the size of the matrix has decreased more than 100 times over several decades. At the same time, new technologies have even increased the sensitivity of one element.
CMOS sensors were invented in the late 1970s. This technology was able to provide higher integration and increased dynamic range. In addition, their production has a lower cost, especially for megapixel sensors. Unfortunately, CMOS sensors had lower sensitivity due to losses in RGB filters. In addition, a large number of transition elements, including amplifiers in the path of each pixel, significantly reduced the useful area of pixels and the uniformity of their parameters. But the active use of innovative technologies has gradually solved most of the problems, providing a rapid growth in the popularity of CMOS matrices. Today they are manufactured by many industry leaders, including Sony, Panasonic, etc.
Pros & Cons
– low production cost and price;
– high performance;
– low power consumption;
– no smearing effect.
CMOS matrices can randomly read each pixel, CCD matrix reads information simultaneously from all pixels. As a result, CMOS matrices do not have the smearing effect that appears in the CCD frame in the form of “light pillar” from bright point objects, for example, the sun or a lantern.
– small area of the photosensitive element relative to the pixel area due to the built-in electronics. As a result, it provides low sensitivity. Unfortunately, the attempt to amplify the signal is accompanied by an increase in image noise;
– “rolling shutter” effect due to line-by-line reading of the signal.
When shooting fast moving objects, sequential reading of lines from top to bottom generates image distortion due to latency.
But today innovative technologies allow companies to create CMOS sensors with support of global shutter mode. These sensors can select between rolling and global shutter modes.
– higher quality global shutter;
– low noise level;
– high sensitivity in the infrared range provides high quality when shooting in low light environments;
– no rolling shutter effect and no vibration sensitivity.
– smearing effect.
In general, CMOS sensors are much cheaper, are convenient for built-in image processing functions, including signal amplification and noise reduction, use simpler firmware with fast tuning and have higher frame rates.
In turn, CCD sensors work as a single unit forming a more organic final image with maximum clarity.
With key features such as vibration stability, good exposure settings and high quality night shooting, more expensive CCDs are commonly used in commercial and racing drones.
Cheaper CMOS sensors are mainly used in budget and mid-budget quadcopters.
In addition, the specs of each camera contain TVL (the number of horizontal lines in the frame), which characterizes its resolution. 300 TVL video corresponds to a very low quality. Today cameras with a matrix of about 600 TVL are the most popular.
Cameras of modern drones support two video formats:
– PAL – 720 × 576 @ 25 fps;
– NTSC – 720 × 480 @ 30 fps.
PAL is handy for medium speed flight and high resolution shooting. NTSC provides smooth shooting during high-speed flights due to minimal delay between frames.
Night cameras are designed for shooting in the infrared range. Some models use very effective and colorful LED backlighting.
In addition, some cameras support the Digital WDR (DWDR) technology for high-contrast scenes. WDR technology creates a balance between too bright and too dark areas of the frame.
Of course, DWDR is inferior to True WDR, which is implemented in hardware.
DWDR software compensates exposure and equalizes color correction. True WDR shoots each frame at a different exposure. In fact, it sets the shutter speed and aperture, providing the optimal brightness range for the sensor with a given sensitivity. Then frames with different aperture exposures are integrated into a single image.
Lenses are selected by thread diameter and focal length. The compact M8 is suitable for nano and micro cameras. M12 are suitable for standard and mini FPV cameras.
The focal length affects the viewing angle during flight.
Of course, modern FPV quadcopters are technically very complex devices. Therefore, choosing the optimal model requires some knowledge of its main components, including the motor, camera, antenna, supported intelligent flight modes, etc.
At the same time, shooting is one of the most demanded functions of a modern drone. Of course, modern intelligent flight modes significantly affect this aspect. But the camera is the second required component.
The choice of cameras directly depends on the pilot level and the use of the quadcopter. Racing drones and commercial models for professional shooting require quite expensive cameras. Therefore, many people prefer the quadrocopters without cameras in order to independently install the required model. The inexpensive onboard CMOS cameras is ideal for novice pilots.
This video offers a comparison between FPV CMOS camera vs CCD camera.