Chinese Optics Letters, 2017, 15 (10): 102801, Published Online: Jul. 19, 2018   

Imaging process and signal-to-noise ratio improvement of enhanced self-heterodyne synthetic aperture imaging ladar Download: 815次

Author Affiliations
1 Key Laboratory of Space Laser Communication and Detection Technology, Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Science, Shanghai 201800, China
2 University of Chinese Academy of Sciences, Beijing 100049, China
Abstract
This Letter gives the general construction of an enhanced self-heterodyne synthetic aperture imaging ladar (SAIL) system, and proposes the principle of image processing. A point target is reconstructed in the enhanced self-heterodyne SAIL as well as in down-looking SAIL experiments, and the achieved imaging resolution of the enhanced self-heterodyne SAIL is analyzed. The signal-to-noise ratio (SNR) of the point target final image in the enhanced self-heterodyne SAIL is higher than that in the down-looking SAIL. The enhanced self-heterodyne SAIL can improve the SNR of the target image in far-distance imaging, with practicality.

Synthetic aperture imaging ladar (SAIL) originates from microwave synthetic aperture radar (SAR). In recent years, side-looking SAIL has demonstrated full two-dimensional (2D) SAIL imaging in the laboratory[17" target="_self" style="display: inline;">7] and airborne[8,9]. There are many inherent defects in side-looking SAIL, such as small optical footprint, limited receiving aperture, need of a matched optical delay, and, in particular, susceptible suffering from the phase interferences of atmospheric turbulence and mechanical trembling in environments. To mitigate the limitations of the side-looking SAIL, a down-looking SAIL was proposed[10]. The down-looking SAIL is based on the key ideas of wavefront transformation of two coaxial and orthogonally polarizaed beams, and a receiver of self-heterodyne detection. The two coaxial and orthogonally polarized beams have the same transmitting and receiving path, thus, in the down-looking SAIL self-heterodynes the two beams could automatically compensate the phase error from atmospheric turbulence and unmodeled line-of-sight motion, and, in principle, no optical delay is needed. Meanwhile, the down-looking SAIL has a controllable and changeable size of the optical footprint, which, together with the associated imaging resolution, can be, manipulated on a large scale. The down-looking SAIL was demonstrated inside laboratory as well as outside under different atmospheric conditions[1113" target="_self" style="display: inline;">–13]. The experiments of airborne down-looking SAIL is reported, and the laser power is 20 W for imaging the target at 3 km[14]. In fact, the laser power is limited and the laser with high power can damage the optical components in the down-looking SAIL transmitting system. The down-looking SAIL uses a self-heterodyne receiver to eliminate phase error. However, the echo signal light is weak at far distances, resulting in low self-heterodyne efficiency and poor quality of the final image. Heterodyne detection is a method of obtaining quantum-limited reception in spectral ranges, and has many advantages in detecting weak signals[15,16].

In this Letter, a enhanced self-heterodyne SAIL based on a down-looking SAIL, is proposed. This system self-heterodynes the two orthogonally polarized echo signals to produce a linear phase history in the orthogonal direction and a quadratic phase history in the travel direction during digital signal processing, and compresses the 2D data in the orthogonal direction by a Fourier transform. Further, it compresses using a match filter with a conjugate quadratic phase of the phase history in the travel direction to get the final target image. The enhanced self-heterodyne SAIL is practical in the far-distance imaging process. The enhanced self-heterodyne SAIL system uses a heterodyne receiver instead of a self-heterodyne receiver. The weak echo signal light interferes with the strong local oscillator, thus the signal-to-noise ratio (SNR) of the target image can be greatly improved.

The structure of the enhanced self-heterodyne SAIL system is illustrated in Fig. 1. The 1×2 polarization-maintaining fiber coupler splits the laser generated by the laser source into two beams, a signal light, and a local oscillator. The signal light is amplified by an optical fiber amplifier and transmitted to the transmitter. The local oscillator is frequency-shifted using the acousto-optic frequency shifter. A rectangular stop near the cylindrical lenses is used to adjust the size of the inner beam. Plano–concave cylindrical lenses L1, L2, L3, and plano–convex cylindrical lens L4 are placed in the focal plane of the transmitting lens. The shift of L1 and L2 produce a linear phase in the orthogonal direction of travel[10].

Fig. 1. Structure of the enhanced self-heterodyne SAIL system. PBS: polarizing beam splitter; WP: wave plate.

下载图片 查看所有图片

The configuration to translate the wavefront using cylindrical lenses L1, L2, L3, and L4 is shown in Fig. 2. Here, cylindrical lenses L1 and L2 move along the orthogonal direction of travel, two cylindrical lenses L3 and L4 move slowly along the direction of travel to produce a spatial quadratic phase history, whereas the target remains stationary.

Fig. 2. Configuration to translate the wavefront using cylindrical lenses; x means the cylindrical lenses L1 and L2 are moving in the direction orthogonal to the direction of travel, y means the cylindrical lenses L3 and L4 are moving in the travel direction.

下载图片 查看所有图片

Based on the structure of the enhanced self-heterodyne SAIL system, the inner optical fields of the two orthogonally polarized beams at the focal plane of the transmitting lens are[1013" target="_self" style="display: inline;">13]eH0(x,y)=C×rectxLxrectyLy×exp{jπλ[(xvxtfSb)2R1+(yvyts)2R1]},eV0(x,y)=C×rectxLxrectyLy×exp{jπλ[(x+vxtfSb)2R1(yvyts)2R2]},where C is the field amplitude of the signal light; λ is the wavelength of the signal light; R1 is the equivalent curvature radius of L1, L2, and L3; R2 is equivalent curvature radius of L4; Lx and Ly are the size of rectangular stops; tf is the fast scanning time; ts is the slow scanning time; vx is the scanning velocity in the x (orthogonal) direction; vy is the scanning velocity in the y (travel) direction; Sb is the distance between the center of the cylinder lens L1 (L2) and the corresponding propagation beam. Using an amplified transmission, the inner field is imaged onto the focal plane of the optical collimator with an amplification factor M, where M=fc/ft. The focal length of the optical collimator is fc, whereas the focal length of the transmitting lens is ft.

The echo signal light is collected from a point target at (xp,yp) by the receiving lens. Then, the optical fields of the two orthogonally polarized echo signal light on the receiving lens are[10]eHr(x,y:xp,yp:tf,ts)=CHexp(jφH),eVr(x,y:xp,yp:tf,ts)=CVexp(jφV),CH=C×rectxpMLxrectypMLy,CV=C×rectxpMLxrectypMLy,φH=πλ[(xpMvxtfMSb)2M2R1+(ypMvyts)2M2R1]+πλfc(xp2+yp2)+πλfc[(xxp)2+(yyp)2]+ΔφH,φV=πλ[(xp+MvxtfMSb)2M2R1(ypMvyts)2M2R2]+πλfc(xp2+yp2)+πλfc[(xxp)2+(yyp)2]+ΔφV.

The phase disturbance caused by the platform vibration and atmospheric turbulence during transmission and reception of the signal light are given by ΔφH and ΔφV, respectively. We assume that there is a uniform local oscillator with constant amplitude. The field of the local oscillator is given by eLO=CLOexp[j(φLO+ωIFt)],where CLO and φLO are the field amplitude and phase of the local oscillator and ωIF is the frequency shift between the signal light and local oscillator.

The echo signal light interferes with the local oscillator, and the resulting signal from the detector can be expressed by a radar equation for point target[10,15]. Hence, we have the signals from the two balanced receivers, UH(xp,yp:tf,ts)=CBDCHLOCHcos[φH+ΔφF(φLO+ωIFt)],UV(xp,yp:tf,ts)=CBDCVLOCVcos[φV+ΔφF(φLO+ωIFt)],where CBD is the photoelectric conversion factor, including all the system constants of integral and photoelectric conversion, also considering the signal light and local oscillator phase front alignment and beam spot size[17]. CHLO and CVLO are the field amplitudes of the local oscillator on the detector. ΔφF is produced when the receiving lens focuses echo signal light onto the detector. During the digital signal processing, to self-heterodyne the two orthogonally polarized-echo signals, UH(xp,yp:tf,ts) and UV(xp,yp:tf,ts) are converted to complex form after the intermediate frequency ωIF is shifted, and UH(xp,yp:tf,ts) is multiplied by the complex conjugate of UV(xp,yp:tf,ts). Thus, we have the 2D data U(xp,yp:tf,ts)=CBD2CHLOCVLOCVCHexp[j(φHφV)]=CBD2CHLOCVLOCVCHexp{jπλ[4(xpMSb)vxtfMR1(ypMvyts)2M2R3]},where 1/R3=1/R1+1/R2. The two coaxial and orthogonally polarized signal lights have the same transmitting and receiving path, thus ΔφHΔφV, so the phase disturbance caused by the platform vibration and atmospheric turbulence can be compensated. During the imaging process, the 2D data U(xp,yp:tf,ts) is first compressed in the orthogonal direction using a Fourier transform. Then, the 2D data is compressed using a match filter with a conjugate quadratic phase of the phase history in the travel direction to get the target image[10].

In our experiment, an optical collimator with a diameter of 600 mm and a focal length of 15 m is designed to simulate the far-distance situation due to the limitations of the laboratory site. A signal light with a wavelength of 1550 nm and a power of 27 dBm is used in the experiment. The local oscillator power is 3 dBm with a 10 kHz frequency shift. The focal length of each of the four cylindrical lenses is 150 mm. In addition, L1, L2, and L3 are plano–concave cylindrical lenses, and L4 is plano–convex cylindrical lens. The scanning velocity of the cylindrical lenses L1 and L2 is 5mm/s, and the scanning time in the orthogonal direction is 1 s, so the scanning range in the orthogonal direction is 5 mm and the corresponding theoretical resolution is 0.29 mm. The scanning velocity of the cylindrical lenses L3 and L4 is 0.003mm/s, and the scanning time in the travel direction is 1600 s, so the synthetic aperture length is 4.8 mm and the corresponding theoretical resolution is 0.30 mm[10]. The transmitting and receiving lenses have a diameter of 200 mm and a focal length of 1200 mm. The dimensions of the rectangular stop Lx and Ly are 7 mm. The balanced amplified photodetector PDB210C has a diameter of 3 mm. The point target is produced by covering the mirror and exposing only an area of 0.3mm×0.3mm and placing it on the focal plane of the optical collimator.

The magnitude and phase distributions of the image focused in the orthogonal direction of travel using the Fourier transform are shown in Figs. 3(a) and 3(b). From Figs. 3(a) and 3(b), the final image resulting from the next focusing in the travel direction using match filtering is shown in Fig. 3(c). The resulting image has a full width at minimum of dx×dy=1.4mm×1.5mm, as shown in Figs. 3(d) and 3(e). The final resolution deviation from the theoretical calculation is mainly due to the nonlinearity of the lens shifting and the phase error of the inner optical fields. The phase error is caused by the cylindrical lens aberration and assembly position error.

Fig. 3. (a) Magnitude and (b) phase distributions of the focused image in the orthogonal direction of travel and (c) a 2D image focused subsequently in the travel direction for a 0.3mm×0.3mm point target. The distribution of the final image is (d) in orthogonal direction and (e) in the travel direction for a 0.3mm×0.3mm point target.

下载图片 查看所有图片

Next, we change the output power of the optical fiber amplifier and image a 0.3mm×0.3mm point target at different output powers. The final image for a 0.3mm×0.3mm point target at output powers of 15 mW, 2 mW, and 750 μW of the optical fiber amplifier are shown in Figs. 4(a)4(c). The distribution of the final image in the travel direction for a 0.3mm×0.3mm point target at output powers of 15 mW, 2 mW, and 750 μW of the optical fiber amplifier are shown in Figs. 4(d)4(f). We calculate the SNR of the point target final image in the travel direction to evaluate the final image quality. The standard deviation of the intensity of all points outside the full width at minimum is the intensity of noise σp. The maximum intensity in the full width at minimum subtracts the average intensity of all points outside the full width is the actual intensity of point target Ip.

Fig. 4. (a), (b), and (c) are the final images for a 0.3mm×0.3mm point target at output power of 15 mW, 2 mW, and 750 μW of the optical fiber amplifier. (d), (e), and (f) are the distribution of the final image in the travel direction for a 0.3mm×0.3mm point target at the output powers of 15 mW, 2 mW, and 750 μW of the optical fiber amplifier.

下载图片 查看所有图片

Thus, the SNR of the point target final image is SNRp=Ip/σp, and the values of SNRp are shown in Table 1.

Table 1. SNR of Point Target Final Image in Enhanced Self-heterodyne SAIL System

Output power of the optical fiber amplifierIntensity of point target IpIntensity of noise σpSNRp
15 mW1127191124
2 mW164115109
750 μW364661

查看所有表

To compare the final image quality of the enhanced self-heterodyne SAIL system and down-looking SAIL, we do not change the transmitter. However, we change the heterodyne receiver to a self-heterodyne receiver by blocking the local oscillator and putting a half-wave plate behind the receiving lens. The structure of the self-heterodyne receiver in a down-looking SAIL is shown in Fig. 5.

Fig. 5. Self-heterodyne receiver in the down-looking SAIL.

下载图片 查看所有图片

Similarly, we change the output power of the optical fiber amplifier and image a 0.3mm×0.3mm point target with the self-heterodyne receiver at different output powers. The final images for a 0.3mm×0.3mm point target in a down-looking SAIL at the output powers of 15 mW, 2 mW, and 750 μW of the optical fiber amplifier are shown in Figs. 6(a)6(c). The distribution of the final image in the travel direction for a 0.3mm×0.3mm point target in a down-looking SAIL at the output powers of 15 mW, 2 mW and, 750 μW of the optical fiber amplifier are shown in Figs. 6(d)6(f). The SNR of the point target final image in a down-looking SAIL are shown in Table 2.

Fig. 6. (a), (b), and (c) are the final images for a 0.3mm×0.3mm point target in the down-looking SAIL at the output powers of 15 mW, 2 mW, and 750 μW of the optical fiber amplifier. (d), (e), and (f) are final images in the travel direction for a 0.3mm×0.3mm point target in the down-looking SAIL at the output powers of 15 mW, 2 mW, and 750 μW of the optical fiber amplifier.

下载图片 查看所有图片

Table 2. SNR of Point Target Final Image in Down-looking SAIL

Output power of the optical fiber amplifierIntensity of point target IpIntensity of noise σpSNRp
15 mW375.283.8797
2 mW61.592.3726
750 μW15.52.466

查看所有表

With an increase in the output power, we can see that the point target final image becomes clear in Figs. 4 and 6, and the SNR of the point target final image has become higher in Tables 1 and 2, which correspond to the theory of the SNR of optical heterodyne detection[15]. Moreover, in the condition of the same output power, it is obvious that the point target final image in Fig. 4 is clearer than that in Fig. 6, and the SNR of the point target final image in Table 1 is higher than that in Table 2. It is clear that an enhanced self-heterodyne SAIL is more advantageous than a down-looking SAIL in far-distance imaging.

In conclusion, we propose the principle of an enhanced self-heterodyne SAIL system. We present SAIL experiments and achieved an imaging resolution of 1.4mm×1.5mm. We compare the SNR of the point target final image in the enhanced self-heterodyne SAIL system and in the down-looking SAIL. The point target final image has a higher SNR and is clearer than that in the down-looking SAIL. The enhanced self-heterodyne SAIL greatly improves the SNR of the target final image, which is practical in far-distance imaging. In our experiment, the power of the local oscillator from the acousto-optic frequency shifter is 3 dBm, the balanced detector does not work in the quantum-limited condition. Thus, the SNR of the heterodyne receiver can be increased by increasing the local oscillator power. However, in this Letter, to compare the SNR of the final image in the enhanced self-heterodyne SAIL and the down-looking SAIL, we specifically image a point target. In is our following work we plan to image a distributed target.

Guo Zhang, Jianfeng Sun, Yu Zhou, Zhiyong Lu, Guangyuan Li, Guangyu Cai, Mengmeng Xu, Bo Zhang, Chenzhe Lao, Hongyu He, Liren Liu. Imaging process and signal-to-noise ratio improvement of enhanced self-heterodyne synthetic aperture imaging ladar[J]. Chinese Optics Letters, 2017, 15(10): 102801.

本文已被 1 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!