Taxonomy and Principles of Distance Measurement

Reflective optical methods are broadly classified into passive and active techniques. Passive methods, such as stereo vision, rely on ambient light to measure distances using two cameras to capture images of a target surface and calculate depth based on relative positions. Active methods, like Time-of-Flight (ToF) cameras, use projected light sources to determine depth by measuring the time it takes for light to reflect off objects and return to the sensor. ToF cameras work by emitting light pulses and analyzing the time delay to calculate distances, while active structured light systems project patterns onto objects to infer depth through deformation analysis. The document also delves into the working principles of ToF technology, including its applications in industrial and consumer electronics, and compares different ToF methods such as Continuous Wave iToF and Pulse iToF, highlighting their respective advantages and limitations.

Reflective optical methods are classified into two main categories: passive and active.

Passive reflective optical methods commonly utilize surrounding light to determine the distance between a target surface and a detector. Stereo vision is an example of such a method, employing two cameras to capture images of a target surface and calculate the distance based on the relative positions of the cameras and the target surface.

A立体视觉系统由两台相机组成，这两台相机安置在同一平面上，通过各自获取同一场景的不同角度图像来构建场景的三维表示。基于两幅图像中捕获到的场景信息，系统通过算法识别图像中的对应点，并利用这些点在两张图像中的位置差异计算场景中物体的相对深度。这一过程即为三角测量。当场景中缺乏足够的几何或颜色特征时，生成的立体图像将呈现均匀状态，此时将缺乏足够的对应像素来通过三角测量获得深度信息。因此，在这种情况下，系统将无法完成立体匹配或深度估计任务。为了有效执行立体视觉任务，场景中应包含丰富的纹理信息或附加人工标记，以便于系统完成相关计算。这些辅助信息为系统的三角测量提供了必要的数据支持。

Active imaging modalities, such as time-of-flight (ToF) cameras and light encoding systems (also known as structured light techniques), do not depend on ambient light. Instead, they utilize a projected light source to acquire three-dimensional range data.

A ToF camera, also known as Time-of-Flight camera, is a category of active depth sensing devices designed to determine the distance of objects by measuring the time taken for light to travel from the camera to the object and back. Through the measurement of time delay, this type of camera is capable of calculating the distance between objects and the camera, and subsequently producing a depth map that represents the spatial distribution of distances.

A structure light system utilizes a projected pattern of light, such as a stripe or checkerboard pattern, to measure the depth of objects within a scene. The pattern is projected onto the scene, and cameras capture the deformation of the pattern as it interacts with the objects. Through analysis of the deformation, depth information is reconstructed to generate a three-dimensional representation of the scene.

Reviewing the Time-of-Flight technology, Vzense has been dedicated to its advancement over the years.

Time-of-Flight (TOF) technology generally covers a range of methods designed to determine the time of flight for a signal traveling from a source to a target and returning. Within the field of photography, TOF cameras specifically employ light-based techniques to measure time of flight, though alternative approaches utilizing diverse signals such as ultrasound waves or electromagnetic radiation also exist. Across all applications, the primary objective remains the calculation of the time taken for a signal to traverse from the source to the target and back, which enables the computation of distance and additional data.

The simplest single-pixel time-of-flight (ToF) technology employs a modulated collimation laser as the transmitter and a single photoelectronic detector as the receiver, enabling the measurement of distance information for a single point. If you aim to utilize a single-pixel distance sensor to obtain the depth map of an entire scene, it typically employs a passive scanning mechanism. The following figure demonstrates the operational principle of single-pixel ToF ranging technology.

该技术通过一次成像实现了完整的场景深度图，无需扫描装置。随着半导体尺寸的不断缩小，该深度相机具备紧凑结构和经济实用的性能，得到了广泛应用和快速发展。

Components of a ToF Camera

The Time-of-Flight (ToF) camera represents a non-scanning 3D depth imaging technology that captures depth information using the optical system as the receiving path. By referring to Figure 1, we can understand that the ToF depth camera is composed of an illuminating unit, an optical lens, an imaging sensor, a control circuit, and a processing circuit.

- Irradiation unit

The irradiation unit is designed to perform pulse modulation on the light source prior to transmission, with a pulse frequency that ranges up to 100MHz. As a result, during the imaging process, the light source is alternately activated and deactivated thousands of times, with each pulse lasting just a few nanoseconds. The exposure time parameter of the camera is governed by the number of pulses captured during each imaging event.

To ensure accurate measurement, precise control of the light pulse is essential, which must have identical duration, rise time, and fall time. Due to even a slight 1ns deviation, the distance measurement error can reach up to 15 cm. Such high modulation frequency and accuracy necessitate the use of highly sophisticated LEDs or laser diodes.

Typically, infrared light sources are not easily detected by human eyes and are employed in various applications.

- Optical lens

This device serves to collect reflected light and capture images on the optical sensor. As opposed to ordinary optical lenses, a band-pass filter is necessary so that only light of the same wavelength as the illumination light source can enter. The main objective of this is to reduce noise interference from uncoordinated light sources and prevent overexposure of the light sensor.

- Imaging sensor

The imaging sensor serves as the central component of the TOF camera. Its structural design is reminiscent of conventional image sensors, yet it is significantly more complex. This sensor incorporates multiple shutters to capture light at various times, enabling precise measurement. Consequently, the pixel size of the TOF chip is notably larger than those found in typical image sensors, usually around 100 micrometers.

- Control unit

The light pulse sequence generated by the camera's electronic control unit is perfectly timed to the operation of the chip's electronic shutter. It acquires and translates sensor charges, which are then transmitted to the analysis unit and data interface.

- Calculation unit

The calculation unit is capable of recording high-precision depth maps. A depth map is typically a grayscale image, with each value indicating the distance from the light reflecting surface to the camera. To achieve better results, data calibration is commonly carried out.

Direct ToF vs Indirect ToF

ToF三维相机技术可按具体实现方式划分为间接式ToF（iToF）和直接式ToF（dToF）。其中，间接式ToF又进一步分为连续波形ToF（Continuous Waveform ToF）和脉冲基ToF（Pulse Based ToF），如图所示：

dToF

The direct time-of-flight (TOF) technique, which constitutes a direct measuring technique for time-of-flight, calculates the time difference between the transmission event (tstart) when the laser pulse is sent from the transmitter and the reception event (tstop) after reflection from an object. Utilizing an internal timer and the known speed of light c, the system determines the distance depth data d. In contrast to the phase-difference-based method, which indirectly measures the time difference between transmission and reception signals, the direct TOF method offers a more straightforward approach, earning it the designation of a direct time-of-flight ranging technique.

The principle of direct time-of-flight ranging is straightforward and easy to implement, yet it demands strict technical specifications for the transmitter's light source, the receiver's image sensor, and the circuits involved in synchronization and time detection. For instance, the transmitter must meet specific requirements to generate extremely short pulses, while the receiver's image sensor must utilize highly sensitive optical detection technology, such as SPAD, to detect weak optical signals.

CW iToF

The fundamental principle of CW iToF is to convert light into a sine wave with a fixed frequency f, with the transmitting end emitting the sine wave at frequency f. Upon receiving the reflected optical energy, the CW iToF system opens multiple observation windows, samples data from each window, analyzes the phase difference information between the transmitted and received signals within a single period, and calculates the distance information using the following formula.

Most continuous waveform ToF systems utilize CMOS sensors, with back-illuminated CMOS technology being particularly prevalent. This approach significantly enhances the light-sensitive area, photon collection efficiency, and ranging speed, with response times as fast as nanoseconds. To achieve phase unwrapping, CW ToF employs multiple modulation frequencies, which proves to be an effective strategy for minimizing multipath errors. As a full CMOS imaging system, CW iToF offers enhanced flexibility and faster readout speeds. However, it also presents certain drawbacks. The image sensor in CW iToF requires the acquisition of four correlation function samples at multiple modulation frequencies, coupled with additional frame processing, which elevates the complexity of signal processing and may necessitate the integration of extra application processors. For measurements involving longer distances, or in scenes with strong ambient light, the system's continuous output power demand becomes substantial, potentially impacting its thermal stability and operational longevity.

Pulse iToF

图1展示了Pulse iTOF工作原理的原理图。通过将光信号调整为固定频率f的方波，使得传输端能够按照频率f发送脉冲信号。接收端的传感器由两个电子快门（S1，S2）组成。S1窗口的频率和相位与发送脉冲保持一致。当S1和S2窗口打开（处于高电平状态）时，它们分别在各自的时间段内积累来自物体的反射光子。通过计算S1和S2窗口中光子能量的比例，分析信号相位以确定发送信号与接收信号的时间差，从而获得距离数据。

Unlike the conventional CW iToF continuous wave debugging mode, the Pulse iToF system offers a more straightforward solution depth, significantly lower computational demands, and enhanced compliance with platform backend capacity constraints. Based on the principle of Pulse iToF, this technology operates by emitting high-intensity light pulses within a brief time window, effectively minimizing the interference of background light signals, thereby enhancing its adaptability to varying ambient light conditions and demonstrating excellent resistance to issues such as scene motion blur. Unlike the conventional CW iToF continuous wave debugging mode, the Pulse iToF system offers a more straightforward solution depth, significantly lower computational demands, and enhanced compliance with platform backend capacity constraints. Based on the principle of Pulse iToF, this technology operates by emitting high-intensity light pulses within a brief time window, effectively minimizing the interference of background light signals, thereby enhancing its adaptability to varying ambient light conditions and demonstrating excellent resistance to issues such as scene motion blur.

Different Technology Comparison