Remote Object Detection Sajjad Haidar, Surrey, BC, Canada Email: [email protected]

Introduction: There are several different ways a remotely (10s of meters or 100s of meters) located object can be detected, or the object’s presence can be ensured. Some of these are namely: 1. Sending ultrasonic sound and receiving the reflected sound from the object. Principle used: Time of Flight measurement or TOF 2. Sending laser or LED light and detecting the reflected light from the object. Principle used: Time of Flight measurement or TOF. 3. Using a CCD camera, comparing successive images. If there is any discrepancy between two images, there could be something wrong with the object. 4. Sending pulsed laser light (Visible or IR), and detecting the scattered light from the object by lock-in amplification. There are other methods as well. Each of the above methods mentioned, have some advantages and disadvantages. Considering several factors I decided to work on a project based on the 4th category. The basic block diagram of the project is shown in Fig.1.

An oscillator (~2.1 kHz) drives a laser diode (LD) through a suitable driver and a LD protection circuit. The pulsed laser light is positioned to shine on a part of the remote object of interest. The faint scattered light is detected by a phototransistor, fitted with a 15 cm long metal pipe (dia. ~5mm). The purpose of this pipe attachment is to make the detection directional. The faint scattered light produces a faint signal at the photo transistor output, which is amplified and filtered (band-pass) to remove unwanted signal and noise. The signal is then applied to one input of a balanced modulator (AD630). The other input of the balanced modulator (also called the reference input) is fed with the delayed original oscillator signal, driving the LD. At the output of the balanced modulator, a DC level is obtained which is proportional to the signal detected at the photo transistor. Any unwanted signal or noise will be eliminated to a large extent by this balanced modulator. This technique of extracting signal also called lock-in-detection. The output is filtered and amplified further. At the final stage a comparator is used to trigger an alarm. When the object is moved or relocated, the signal goes below a certain threshold level and the alarm sounds to alert.

Fig.1. System block diagram

Oscillator and LD Driver: The oscillator used here is based on simple 555 timer IC in astable operation. The circuit diagram is shown in Fig.2. Using a 50 K potentiometer (R8), frequency is set at ~ 2.1 kHz. The duty cycle is ~ 50%. The output of the oscillator is fed to a MOSFET via a buffer (OPA2171), which is driving the LD. The LD used here, is a cheap laser pointer. The laser pointer has an LD as well as a focusing lens in front of it. While using an LD in pulsed operation, we have to consider its protection as well. Resistor R2 (87 ) is used to limit the current within the safe level of ~ 40 mA. A ferrite bead (FB) in series and a ceramic capacitor (0.1F) in parallel to the LD is used to achieve slow rise and fall time of the current through the LD. As mentioned earlier, oscillator output will also be used as a reference input for the lock-in-amplification. The buffered oscillator output cannot be effectively used directly for this purpose. Because of the LD protection circuit there will be a little delay in turning ON and OFF the LD; also there will be delay in sending the pulses and receiving the signal also known as time of flight (TOF). At the receiving end, the phototransistor detector output is amplified and filtered, these circuits also introduced some delay. To match with the delayed pulses the reference signal also needs to be

delayed. The pulse delay circuit is implemented by an IC, LTC6994. The amount of delay can be adjusted by a 150K pot (R9). Before being fed to the reference input, the DC component of the pulses is removed by a capacitor C3. The implemented PCB is also shown in Fig.3. The oscillator output or the LD driver input and the delayed output waveforms are shown in Fig.4 and Fig.5, respectively.

Fig.2. Oscillator, LD driver and delay circuit

Fig.3. PCB implementation of the oscillator, LD driver and delay circuit

Fig.4. Oscillator and the delayed reference output waveforms

Fig.5. Voltage across the LD

Phototransistor Amplifier and Filter: The pulsed laser light is incident on the object to be detected and only a tiny fraction of the scattered light reaches back to the detection system. This scattered light is detected by a phototransistor. The phototransistor is fitted to a 5 mm diameter and 15 cm long metal tube to make directional sensing possible. The phototransistor current is amplified by an op amp (OPA21180). Along with the pulsed scattered light from the object, ambient light or background light also generates photo-current. 60 Hz line hum is also present at the amplifier output to a little extent. To remove those unwanted signal and background noise a simple notch filter is used. The passive components are chosen to pass only the pulsed light signal at 2.1 kHz. Other signals and noises are heavily attenuated. The phototransistor and the amplifier-filter is implemented in the same PCB, and put in a small enclosure. The cable from this closure carries photo-signal to the lock-in amplifier and power also comes to it from the common power supply. The schematic and the implemented PCB is shown in Fig. 6 and Fig.7, respectively.

Fig.6. Phototransistor amplifier and band pass filter.

Fig.7. Phototransistor amplifier PCB, and the enclosure.

Lock-in Detection Amplifier: The heart of the detecting system is a balanced modulator, implemented by IC-AD630 and is shown in Fig.8. AD630 has two inputs, we can name one a signal input and other a reference input. The signal obtained from the photo-detector is amplified by IC1A (OPA2171A) and applied to input RINA. The delayed reference signal is applied to the other input SEL. We can either apply to SEL_A or to SEL_B, depending on the polarity of the reference signal. Either of these two inputs is selected by two jumpers. The polarity of the detected output voltage depends on these jumper selectors. As the expected output is of positive output voltage, we can change the jumper position, if we find negative output.

Fig.8. The schematic of the lock-in amplifier The balanced modulator (AD630) actually multiplies the two input signals and the multiplied result is produced at the output. Only the two signals of exact frequency and phase will produce a DC output, others will cancel out. That is why lock-in detection is a good technique to detect only our desired signal, though very weak in nature. As the output is a pulsating DC, a passive RC filter consisting of R7 and C9 is used to obtain stable DC, it also filters out any small and fast variation of the desired DC output. As the scattered light produces very little signal, the resulting demodulated DC output is also very low, a DC amplifier consisting of IC3A and IC3B lifts this to usable level of several hundred millivolts to few volts. Off course the DC output depends on the distance of the object as well. The further the object the less is the output DC level. This output DC can be seen on a DC voltmeter or on an

oscilloscope. The oscilloscope-waveforms of the signal input, reference input, demodulated output and amplified DC output is shown in Fig. 10.

Fig.9. The implemented PCB of the lock-in amplifier and the associated input-output terminals and control knobs. To detect the presence and absence of the distant object, it is better to have some sharp indication (ON or OFF) as well along with the DC level measurements. Or in many occasions the sharp detection will suffice. For this purpose a comparator (IC4A) is used at the output of the DC amplifier stage. We can set a DC reference level by a pot in the front panel. Upon receiving a small scattered light by the phototransistor and going through the all the processes, a DC voltage proportional to the received light is applied at the non-inverting terminal of the comparator. If the front-panel pot (R11) is now adjusted to little lower than this voltage, the comparator output will be HIGH. If the distant object is removed, the DC signal drops and the comparator output will go LOW. The comparator output can also be used to turn ON an alarm by employing an electro-mechanical or solid state relay.

Fig.10. Signal waveforms at various test points of the lock-in amplifier The photograph of the entire distant object detection system is shown in Fig.11. As it is a prototype one, the units are separated for easy testing. The final product could be reduced to one or two units. In this prototype version a camera stand is used to hold the small pipe fitted with the photo transistor. Another pipe of similar size is attached just onto the first one, which serves as a kind of object-sight using naked eye. Targeting the laser-lighted object through the pipe-hole will give an output signal at TP4, little adjusting the horizontal and azimuth of the tripod, maximizes the signal output. Now controlling the comparator-threshold just below the output level we can make the comparator-output zero, If the object is moved

Fig.11. The photograph of the remote object detection system.

Bill of Material (BOM): Phototransistor Amplifier and Filter:

Qty.

Value

Device

Package

S04P

1

Parts

Description

S04P

CON

amp mt connector

1

100 k

R0805

R0805

R1

resistor

2

10 nF

C0805

C0805

C2, C3

capacitor

2

10 F 120 k

CPOL-SMCB

SMC_B

C4, C5

polarized capacitor

R0805

R0805

R3

resistor

R0805

R0805

R5

resistor

1

365  500 pF

C0805

C0805

C1

capacitor

2

6.8 k

R0805

R0805

R2, R4

resistor

1

BPW77NA

BP103

BP103

T1

photo transistor

1

OPA21180

OPA2111D

SO08

IC1

op amp

1 1

Lock-in Amplifier Qty.

Value

Device

Package Parts

2

Connector

03P

4

Connector

S02P

2

Connector

S03P

Description

JMP1, JMP2 OUT1, OUT2, REF1, SIGIN POT, PWR C1, C2, C3, C4, C5, C7, C10, C11, C12, C13, C14, C15 C6, C8

amp quick connector

R1, R3, R5, R8, R12

resistor

C9

polarized capacitor

amp mt connector amp mt connector

12

0.1 F

C1206K

C1206K

2

100 F 10 k

CPOL-E5-6

E5-6

M1206

M1206 E5-6

3

10 F 1k

CPOL-E5-6 M1206

M1206

R6, R7, R9

resistor

1

20 k

M1206

M1206

R2

resistor

1

33 k

M1206

M1206

R4

resistor

1

5k

Potentiometer

B25U

R11

potentiometer

1 AD630ARZ

AD630ARZ

SOIC12

IC2

balanced mod./demod.

3 OPA2171A

LM358D

SO08

IC1, IC3, IC4

op amp

PTR1PAD1-13

P1-13

TP2, TP3, TP4

test pin

5 1

3

Test-Pin

capacitor polarized capacitor

Oscillator and LD Driver: Qty.

Value

3

Device

Package

Parts

Description

Connector

S02P

I/O

amp mt connector

1

0.01 F

C-050

C050-025X075

C5

capacitor

4

0.1 F 100 k

C-050

C050-025X075

C1, C4, C7

capacitor

1

0207/10

0207/10

R10

resistor

E5-6

C6

polarized capacitor

2

100 F 10 k

CPOL-E5-6 0207/10

0207/10

R3, R6

resistor

1

10 F

C-050

C050-025X075

C2

capacitor

1

0207/10

0207/10

R1

resistor

1

150  150 k

TRIM -B64W

B64W

R9

potentiometer

1

1k

0207/10

0207/10

R4

resistor

1

C-050

C050-025X075

C3

capacitor

1

1 F 2k

0207/10

0207/10

R7

resistor

1

3k

0207/10

0207/10

R5

resistor

1

50 k

TRIM -B64W

B64W

R8

potentiometer

1

87  IRL2703

0207/10

0207/10

R2

resistor

IRF540

TO220BV

Q1

MOSFET

LTC6994CS6

SOT6

IC3

delay block

NE555D

NE555D

SO08

IC1

timer

1 OPA2171

LM358D

SO08

IC2

op amp

1

1

1 LTC6994 1

Remote Object Detection

and C9 is used to obtain stable DC, it also filters out any small and fast variation of the .... I/O amp mt connector. 1 0.01 μF. C-050. C050-025X075 C5 capacitor.

1MB Sizes 24 Downloads 365 Views

Recommend Documents

Scalable High Quality Object Detection
Dec 9, 2015 - posal quality over its predecessor Multibox [4] method: AP increases from 0.42 to ... call compared to Multiscale Combinatorial Grouping [18] with less proposals on the ... proposal ranking function provides a way to balance recall ....

Semantic-Shift for Unsupervised Object Detection - CiteSeerX
notated images for constructing a supervised image under- standing system. .... the same way as in learning, but now keeping the factors. P(wj|zk) ... sponds to the foreground object as zFG and call it the fore- ..... In European Conference on.

Detection-based Object Labeling in 3D Scenes
In computer vision, both scene labeling and object de- tection have been extensively .... Hence, the maximum detector response across all scales is the best estimate of the true ..... at 30, 45, and 60 degrees with the horizon. Each detector is.

Scalable Object Detection using Deep Neural Networks
neural network model for detection, which predicts a set of class-agnostic ... way, can be scored using top-down feedback [17, 2, 4]. Us- ing the same .... We call the usage of priors for matching ..... In Proceedings of the IEEE Conference on.

Salient Object Detection by Composition
mining an appropriate context in advance is difficult and incorrect .... ground. In [25], low spatial variance of a feature is con- ..... data using bidirectional similarity.

Object Detection by Compressive Sensing
[4] demonstrate that the most discriminative features can be learned online to ... E Rn×m where rij ~N(0,1), as used in numerous works recently [9]. ..... 7.1 Sushma MB has received Bachelor degree in Electronics and communication in 2001 ...

Deep Neural Networks for Object Detection - NIPS Proceedings
This method combines a set of discriminatively trained .... network to predict the object box mask and four additional networks to predict four ... In order to complete the detection process, we need to estimate a set of bounding ... training data.

Object Detection and Viewpoint Estimation with Auto ...
viewpoint ground truth and more information about object appearance in train- ... such as pedestrian detection [21], face verification [9], face parsing [16], and.

Object Detection in Video with Graphical Models
demonstrate the importance of temporal information, we ap- ply graphical models to the task of text detection in video and compare the result of with and without ...

Hierarchical Co-salient Object Detection via Color Names - GitHub
HCN. Figure 2: Illustration of single-layer combination. Border Effect. In the testing data set, some of the input images have thin artificial borders, which may affect.

Integrating Visual Context and Object Detection within a ...
other promising papers on visual context for object detection [10,8,11,9], we define the .... The LabelMe image database [13] is often used for such pur- poses.

Active Skeleton for Non-rigid Object Detection
results on two widely used benchmark image sets: the Weiz- .... In the detection stage, we avoid the difficult steps of per- ... (1) skeleton branches represent meaningful parts of an ob- ... posed in [23] as the distance measure for D(Eg(I),Fi(Θ)).

Shape Band: A Deformable Object Detection Approach
Huazhong Univ. of Science and Technology ... Dept. of Neurology and Dept. of Computer Science ... resentation, which has the tolerance for certain degree of.