Traffic surveillance system

• US 5,161,107 A
• Filed: 10/25/1990
• Issued: 11/03/1992
• Est. Priority Date: 10/25/1990
• Status: Expired due to Fees

• Current Assignee: Mestech Creation Corporation
• Original Assignee: Mestech Creation Corporation
• Inventors: Mayeaux, Paul H., Stamps, Virgil O.
• Primary Examiner(s): Black, Thomas G.
• Application Number:
  US07/603,179
• Time in Patent Office: 740 Days
• Field of Search: 340/937, 364/436, 364/424.01, 382/18
• US Class Current: 701/117
• CPC Class Codes: G06V 10/143 Sensing or illuminating at different wavelengths
• CPC Class Codes: G06V 20/54 of traffic, e.g. cars on the road, trains or boats
• CPC Class Codes: G08G 1/04 using optical or ultrasonic detectors
• CPC Class Codes: H04N 23/20 for generating image signals from infrared radiation only
• CPC Class Codes: H04N 23/667 Camera operation mode switching, e.g. between still and video, sport and normal or high- and low-resolution modes
• CPC Class Codes: H04N 5/33 Transforming infrared radiation cameras or camera modules for generating image signals from infrared radiation H04N23/20; circuitry of SSIS for transforming infrared radiation into image signals H04N25/20

Abstract

A machine vision system for detecting and recognizing objects. The system includes a camera which forms three different images of the object using different portions of the electromagnetic spectrum and uses these different images depending on the time of day and weather conditions. The machine vision system also enables an operator to select zones within the camera'"'"'s field of view than may be of any size or shape. The machine vision system only detects objects in the specified zones. The vision system provides signals to a traffic light controller to provide vehicle or object presence indications for traffic signal sequencing purposes. The vision system can simply detect an object or optionally can pattern match for specific objects.

Claims

1. A traffic surveillance system for detecting the presence of vehicles and for generating signals for operating a traffic light, comprising:

means for receiving electromagnetic wavelengths of varying frequencies from the vehicle;

a plurality of filter means for receiving said electromagnetic wavelengths from said receiving means and passing portions of said electromagnetic wavelengths according to said varying frequencies in the electromagnetic spectrum, each filter means passing a different portion of the electromagnetic spectrum;

means receiving the electromagnetic wavelengths passed by said plurality of filter means for translating said portions of said electromagnetic wavelengths passed by said filter means into a corresponding plurality of images;

means coupled to said translating means for selecting one of said images formed from said portions of said electromagnetic wavelength as a function of varying detectability conditions;

image processing means coupled to said selecting means which uses said selected image to determine if a vehicle is present and generates a signal indicative thereof; and

means coupled to said image processing means for receiving said indicating signal and for changing the traffic light according to said signal.

2. The traffic surveillance system of claim 1, wherein said selecting means includes:

clocking means for differentiating between daytime and nighttime, and

wherein said selecting means selects a default image from, a first portion of said electromagnetic wavelengths during daytime and a default image from a second portion of said electromagnetic wavelengths during nighttime.

3. The traffic surveillance system of claim 2, wherein said selecting means further includes:

clocking means coupled to said image processing means for measuring a first length of time when no vehicle is present; and

means coupled to said clocking means for changing to an image other than said default image when said first length of time has expired.

4. The traffic surveillance system of claim 3, wherein said selecting means further includes:

clocking means coupled to said image processing means for measuring a second length of time that no vehicle is present when said selecting means has selected said non-default image; and

means coupled to said clocking means for returning to said default image when said second length of time has expired.

5. The traffic surveillance system of claim 4, wherein said first length of time is greater than said second length of time.

6. The traffic surveillance system of claim 2, wherein said image processing means includes a means for generating edge density values from said default images;

and

wherein said selecting means includes;

means for generating a running average of said default image edge density value;

means coupled to said generating means for comparing said running average edge density value with said default image edge density value; and

means coupled to said comparing means for selecting an image other than said default image when said default image edge density value is less than said running average edge density value by a predetermined amount.

7. The traffic surveillance system of claim 1, wherein each of said filter means comprises a bandpass filter.

8. The traffic surveillance system of claim 7, wherein said bandpass filters include bandpass filters operating at the visible light spectrum, the ultraviolet spectrum, and the infrared spectrum.

9. A machine vision system for detecting and recognizing objects, comprising:

means for receiving electromagnetic radiation from the object;

a plurality of filter means receiving said electromagnetic radiation from said receiving means and passing portions of said electromagnetic radiation according to frequencies in the electromagnetic spectrum, each filter means passing a different portion of the electromagnetic spectrum;

means for translating said portions of said electromagnetic radiation passed by said filter means into a corresponding plurality of images;

storing means for storing reference object images; and

image processing means coupled to said translating means which compares portions of one of said images with said reference object images to determine if an object is present and generates signals indicative of which of said reference object images most closely matches said portion of said image.

10. A traffic surveillance system for detecting the presence of vehicles, comprising:

means for receiving electromagnetic wavelengths of varying frequencies from the vehicle, wherein said receiving means has a field of view;

image processing means coupled to said translating means to determine if a vehicle is present in said field of view, said image processing means including;

(a) means for converting said image into a digital image comprising a plurality of pixels having associated values;

(b) latching means for storing a threshold value;

(c) comparator means coupled to said converting means and said latching means for comparing said threshold value with said associated values of each of said pixels and generating a signal value indicating thereof;

(d) means coupled to said comparator means for storing each of said comparator means signal value for each pixel; and

(e) means coupled to said storing means for receiving said stored comparator means signal values and generating an image edge density value representing the percentage of said comparator means signal values greater than said threshold value.

11. The traffic surveillance system of claim 10, wherein said image processing means further includes:

means for receiving said image edge density valve and for determining information therefrom regarding the presence of a vehicle in said field of view.

12. The traffic surveillance system of claim 11, further comprising selection means for selecting said threshold value for each of multiple zones in said field of view.

13. The traffic surveillance system of claim 12, further comprising:

control means for selecting a rectangular-shaped zone for each of said multiple zones in said field of view.

14. The traffic surveillance system of claim 11, wherein said comparator means signal has a first value if said pixel value if greater than said threshold value and said comparator means signal has a second value if said pixel value is less than said threshold value;

and

wherein said image processing means further includes;

background means coupled to said storing means for generating an edge density value representing the percentage of said comparator means signal values having said first value when no vehicle is present in said field of view;

subtracting means coupled to said image edge density value means and said background means for subtracting said background edge density value from said image edge density value and generating an object edge density value; and

comparing means coupled to said subtracting means for comparing said object edge density value with a limit value and generating a signal indicative thereof.

15. The traffic surveillance system of claim 14, further comprising selecting means for selecting said limit value for each of multiple zones in said field of view.

16. A machine vision system for detecting the presence of objects, comprising:

means for receiving electromagnetic wavelengths of varying frequencies from the object;

a plurality of filter means for receiving said electromagnetic wavelengths from said receiving means and passing portions of said electromagnetic wavelengths according to said varying frequencies in the electromagnetic spectrum, each filter means passing a different portion of the electromagnetic spectrum;

means receiving the electromagnetic wavelengts passed by said plurality of filter means for translating said portions of said electromagnetic radiation passed by said filter means into a corresponding plurality of images;

means coupled to said translating means for selecting one of said images formed from said portions of said electromagnetic wavelengths as a function of varying detectability conditions; and

image processing means coupled to said selecting means which uses said selected image to determine if an object is present and generates a signal indicative thereof.

17. The machine vision system of claim 16, wherein said selecting means includes:

clocking means for differentiating between daytime and nighttime, and

wherein said selecting means selects a default image from a first portion of said electromagnetic wavelength during daytime and a default image from a second portion of said electromagnetic wavelength during nighttime.

18. The machine vision system of claim 17, wherein said selecting means further includes:

clocking means coupled to said image processing means for measuring a first length of time that no object is present; and

means coupled to said clocking means for changing to an image other than said default image when said first length of time has expired.

19. The machine vision system of claim 18, wherein said selecting means further includes:

clocking means coupled to said image processing means for measuring a second length of time that no object is present when said selecting means has selected said non-default image; and

means coupled to said clocking means for returning to said default image when said second length of time has expired.

20. The machine vision system of claim 19, wherein said first length of time is greater than said second length of time.

21. The machine vision system of claim 17, wherein said image processing means includes a means for generating edge density values from said default images;

and

wherein said selecting means includes;

means for generating a running average of said default image edge density value;

means coupled to said generating means for comparing said running average edge density value with said default image edge density value; and

means coupled to said comparing means for selecting an image other than said default image when said default image edge density value is less than said running average edge density value by a predetermined amount.

22. The machine vision system of claim 16, wherein each of said filter means comprises a bandpass filter.

23. The machine vision system of claim 22, wherein said bandpass filters include bandpass filters operating at the visible light spectrum, the ultraviolet spectrum, and the infrared spectrum.

24. A machine vision system for detecting and recognizing objects, comprising:

means for receiving electromagnetic wavelengths of varying frequencies from the object;

a plurality of filters receiving said electromagnetic wavelengths from said receiving means and each passing a portion of said electromagnetic wavelengths as a function of the frequency of the electromagnetic spectrum;

means for translating portions of said electromagnetic wavelengths into a corresponding plurality of images;

storing means for storing multiple reference object images; and

imaging processing means coupled to said translating means which compares at least one of said plurality of images with said reference object images to determine if an object is present and generates signals indicative of which of said multiple reference object images most closely matches said at least one of said plurality of images.

25. The machine vision system of claim 24, further comprising:

selecting means for selecting one of the plurality of images passed through a respective one of the filters as a function of varying detectability conditions.

26. The machine vision system of claim 25, wherein the selecting means comprising:

clocking means for differentiating between daytime and nighttime; and

wherein said selecting means selects a default image from a first portion of said electromagnetic wavelengths during daytime and a default image from a second portion of said electromagnetic wavelengths during nighttime.

27. The machine vision system of claim 26, wherein said selecting means further includes:

clocking means coupled to said image processing means for measuring a first length of time when no vehicle is present; and

means coupled to said clocking means for changing to an image other than said default image when said first length of time has expired.

28. The machine vision system of claim 27, wherein said selecting means further includes:

clocking means coupled to said image processing means for measuring a second length of time when no vehicle is present when said selecting means has selected said non-default image; and

means coupled to said clocking means for returning to said default image when said second length of time has expired.

29. The machine vision system of claim 28, wherein said first length of time is greater than said second length of time.

30. The machine vision system of claim 26, further comprising:

said image processing means includes a means for generating edge density value from said default images; and

wherein said selecting means includes;

means for generating a running average of said default image edge density value;

means coupled to said generating means for comparing said running average edge density value with said default image edge density value; and

means coupled to said comparing means for selecting an image other than said default image when said default image edge density value is less than said running average edge density value by a predetermined amount.

31. The machine vision system of claim 24, further comprising:

said means for receiving electromagnetic wavelengths has a field of view; and

control means for selecting a plurality of zones within said field of view, each of said plurality of zones being a rectangular-shaped zone.