Table 2.

Advantages and disadvantages of the most common sensor implementations, based on the basic principles for 3-D vision.

Basic Principle	Sensor/Technique	Advantages	Disadvantages
Triangulation	Consumer triangulation sensor (CTS)	-Off-the-shelf -Low cost -Provide RGB stream -Good community support, good documentation -Open source libraries available	-Vulnerable to sunlight, where no depth information is produced -Depth information is not possible at night or in very dark environments -Not weather resistant -Warm-up time required to stabilize the depth measurements (~1 h)
	Stereo vision	-Good community support, good documentation -Off-the-shelf smart cameras (with parallel computing) available -Robust enough for open field applications	-Low texture produce correspondence problems -Susceptible to direct sunlight -Computationally expensive -Depth range is highly dependent on the baseline distance
	Structure-from-motion	-Digital cameras are easily and economically available -Open source and commercial software for 3-D reconstruction -Suitable for aerial applications -Excellent portability	-Camera calibration and field references are a requirement for reliable measurements -Time consuming point cloud generation process is not suitable for real-time applications -Requires a lot of experience for obtaining good raw data
	Light sheet triangulation	-High precision -Fast image data acquisition and 3-D reconstruction -Limited working range due to the focus -Do not depend on external light sources -New versions have light filtering systems that allow them to handle sunlight	-High cost -Susceptible to sunlight -Time consuming data acquisition
TOF	TOF camera	-Active illumination independent of an external lighting source -Able to acquire data at night or in dark/low light conditions -Commercial 3-D sensors in agriculture are based on the fast-improving photonic mixer device (PMD) technology -New versions have pixel resolutions of up to 4.2 Megapixels -New versions have depth measurement ranges of up to 25 m	-Most of them have low pixel resolution -Most of them are susceptible to direct sunlight -High cost
	Light sheet (pulse modulated) LIDAR	-Emitted light beams and are robust against sunlight -Able to retrieve depth measurements at night or in dark environments -Robust against interference -Widely used in agricultural applications -Many research papers and information available -New versions perform well in adverse weather conditions (rain, snow, mist and dust)	-Poor performance in edge detection due the spacing between the light beams -Warm-up time required to stabilize the depth measurements (up to 2.5 h) -Normally bulky and with moving parts -Have problems under adverse weather conditions (rain, snow, mist and dust)
Interferometry	Optical coherent tomography (OCT)	-High accuracy -Near surface light penetration -High resolution	-High cost -Limited range -Highly-textured surfaces scatter the light beams -Relative measurements -Sensitive to vibrations -Difficult to implement