Method and system for detecting an object using a composite evidence grid

• US 7,142,150 B2
• Filed: 12/15/2004
• Issued: 11/28/2006
• Est. Priority Date: 12/15/2004
• Status: Active Grant

• Current Assignee: Deere & Company
• Original Assignee: Deere & Company
• Inventors: Thackray, Robert George
• Primary Examiner(s): Gregory, Bernarr E.
• Application Number:
  US11/013,243
• Publication Number:
  US 20060125680A1
• Time in Patent Office: 713 Days
• Field of Search: 342/27, 342/28, 342/52, 342/54, 342 70- 72, 342 89-103, 342/118, 342128-146, 342/175, 342/195, 342147-158, 701/300, 701/301, 180167-169
• US Class Current: 342/54
• CPC Class Codes: G01S 13/865 Combination of radar systems with lidar systems
• CPC Class Codes: G01S 13/88 Radar or analogous systems specially adapted for specific applications electromagnetic prospecting or detecting of objects, e.g. near-field detection, G01V3/00
• CPC Class Codes: G01S 13/931 of land vehicles
• CPC Class Codes: G01S 17/86 Combinations of lidar systems with systems other than lidar, radar or sonar, e.g. with direction finders
• CPC Class Codes: G01S 17/88 Lidar systems specially adapted for specific applications
• CPC Class Codes: G01S 17/931 of land vehicles
• CPC Class Codes: G01S 7/40 Means for monitoring or calibrating
• CPC Class Codes: G01S 7/41 using analysis of echo signal for target characterisation; Target signature; Target cross-section
• CPC Class Codes: G01S 7/4802 using analysis of echo signal for target characterisation; Target signature; Target cross-section
• CPC Class Codes: G01S 7/497 Means for monitoring or calibrating

Abstract

A method and system for detecting an object uses a composite evidence grid based on dual frequency sensing. A source transmits a laser transmission in a first zone. A detector receives a reflection of the laser transmission from an object in the first zone to determine laser observed data associated with points on the object. A transmitter transmits a radar transmission in a second zone that overlaps with the first zone. A receiver receives a reflection of the radar transmission from an object in the second zone to determine radar observed data associated with points on the object. The laser observed data is processed to form a laser occupancy grid for the first zone and the radar observed data is processed to form a radar occupancy grid for the second zone. An evaluator evaluates the radar occupancy grid and the laser occupancy grid to produce a composite evidence grid for at least an overlapping region defined by the first zone and the second zone.

Claims

1. A method of detecting an object, the method comprising:

transmitting a laser transmission over a first frequency range in a first zone;

receiving a reflection of the laser transmission from an object in the first zone to determine laser observed data associated with points on the object;

processing the laser observed data to form a later occupancy grid for the first zone;

transmitting a radar transmission over a second frequency range in a second zone that overlaps with the first zone;

receiving a reflection of the radar transmission from an object in the second zone to determine radar observed data associated with points on the object;

processing the radar observed data to form a radar occupancy grid for the second zone; and

evaluating the radar occupancy grid and the laser occupancy grid to produce a composite evidence grid for at least an overlapping region defined by the first zone and the second zone.

2. The method according to claim 1 further comprising:

applying a first data processing technique in accordance with Bayes law to the observed laser data associated with different corresponding sample times to determine a laser occupancy grid; and

applying a second data processing technique in accordance with Bayes law to the observed radar data associated with different corresponding sample times to determine a temporal radar occupancy grid.

3. The method according to claim 2 wherein the evaluating comprises applying a third data processing technique in accordance with Bayes law with respect to the laser occupancy grid and the radar occupancy grid to form a composite evidence grid for an overlapping region defined by an intersection of the first zone and the second zone.

4. The method according to claim 1 wherein the laser occupancy grid is a two-dimensional grid;

each cell in the laser occupancy grid containing a probability with a range from 0 to 1 that a respective spatial region represented by the cell of the grid is occupied by an object; and

wherein the radar occupancy grid is a two-dimensional grid;

each cell in the radar occupancy grid containing a probability with a range from 0 to 1 that a respective spatial region represented by the cell is occupied by an object.

5. The method according to claim 1 wherein the laser occupancy grid is a three-dimensional grid;

each cell in the laser occupancy grid containing a probability with a range from 0 to 1 that a respective spatial region in the grid is occupied by an object and;

wherein the radar occupancy grid is a three-dimensional grid;

each cell in the radar occupancy grid containing a probability with a range from 0 to 1 that a respective spatial region in the grid is occupied by an object.

6. The method according to claim 1 further comprising transforming probability levels of the cells of the laser occupancy grid for a non-overlapping region within the first zone to form an addendum evidence grid;

and combining the addendum evidence grid to the composite evidence grid to form an augmented evidence grid.

7. The method according to claim 1 further comprising:

defining an object representation for the object in the first zone by defining a boundary of the object from the composite evidence grid and estimating at least one dimension associated with the defined boundary.

8. The method according to claim 7 further comprising:

characterizing the object as a trunk, a limb, or another tree portion based on the defined object representation.

9. A system for detecting an object, the system comprising:

a source for transmitting a laser transmission in a first zone;

a detector for receiving a reflection of the laser transmission from an object in the first zone to determine laser observed data associated with points on the object;

a transmitter for transmitting a radar transmission in a second zone that overlaps with the first zone;

a receiver for receiving a reflection of the radar transmission from an object in the second zone to determine radar observed data associated with points on the object;

a data processing system for processing the determined laser observed data to form a laser occupancy grid for the first zone and for processing the determined radar observed data to form a radar occupancy grid for the second zone; and

an evaluator for evaluating the radar occupancy grid and the laser occupancy grid to produce a composite evidence grid for at least an overlapping region defined by the first zone and the second zone.

10. The system according to claim 9 wherein the data processing system applies a first data processing technique in accordance with Bayes law to the observed laser data associated with different corresponding sample times to determine a laser occupancy grid;

the data processing system applies a second data processing technique in accordance with Bayes law to the observed radar data associated with different corresponding sample times to determine the radar occupancy grid.

11. The system according to claim 10 wherein the evaluator applies a third data processing technique in accordance with Bayes law with respect to the laser occupancy grid and the radar occupancy grid to form a composite evidence grid for the overlapping region defined by the first zone and the second zone.

12. The system according to claim 9 wherein the laser occupancy grid is a two-dimensional grid;

each cell in the laser occupancy grid containing a probability with a range from 0 to 1 that a respective spatial region represented by the cell of the grid is occupied by an object; and

wherein the radar occupancy grid is a two-dimensional grid;

each cell in the radar occupancy grid containing a probability with a range from 0 to 1 that a respective spatial region represented by the cell is occupied by an object.

13. The system according to claim 9 wherein the laser occupancy grid is a three-dimensional grid;

each cell in the laser occupancy grid containing a probability with a range from 0 to 1 that a respective spatial region in the grid is occupied by an object and;

wherein the radar occupancy grid is a three-dimensional grid;

each cell in the radar occupancy grid containing a probability with a range from 0 to 1 that a respective spatial region in the grid is occupied by an object.

14. The system according to claim 9 further comprising:

a definer for defining an object representation for the object in the first zone and the object in the second zone, where the object represents a single object.

15. The system according to claim 14 further comprising:

a characterization module for characterizing the object as a trunk, a limb, or another tree portion based on the defined object representation.

16. The system according to claim 9 further comprising:

a definer for defining an object representation for the object in the first zone;

the definer comprising a boundary definition module for defining a boundary of the object from the composite evidence grid and a dimension estimator for estimating at least one dimension associated with the defined boundary.

17. A method of detecting an object, the method comprising:

transmitting a laser transmission at a known angular orientation over a first frequency range in a first zone;

receiving a reflection of the laser transmission from an object in the first zone to determine laser range data associated with points on the object for known angular orientation data;

processing the laser observed data to form a laser occupancy grid for the first zone;

transmitting a radar transmission at a known angular position over a second frequency range in a second zone that overlaps with the first zone;

receiving a reflection of the radar transmission from an object in the second zone to determine radar range data associated with points on the object for the known angular position data;

processing the radar range data and corresponding known angular position data to form a radar occupancy grid for the second tone; and

evaluating the radar occupancy grid and the laser occupancy grid to produce a composite evidence grid for at least an overlapping region defined by the first zone and the second zone.

18. The method according to claim 17 further comprising:

applying a first data processing technique in accordance with Bayes law to the observed laser data associated with different corresponding sample times to determine a laser occupancy grid; and

applying a second data processing technique in accordance with Bayes law to the observed radar data associated with different corresponding sample times to determine a temporal radar occupancy grid.

19. The method according to claim 18 wherein the probability levels of the laser occupancy grid are determined in accordance with the following equation:

$p<br><br>\text{〈<br><br>}<br><br>l_{i,j}<br><br><br><br>m_{t,\mathrm{\dots <br><br>}<br><br>,}<br><br>m_1\rangle <br><br>=\frac{p<br><br>\langle <br><br>m_t<br><br><br><br>l_{i,j}\rangle <br><br><br><br>p<br><br>\langle <br><br>l_{i,j}<br><br><br><br>m_{t-1,\mathrm{\dots <br><br>}<br><br>,}<br><br>m_0\rangle <br><br>}{p<br><br>\left(m_t\right)},\mathrm{where}<br><br><br><br><br><br>p<br><br>\langle <br><br>l_{i,j}<br><br><br><br>m_{t,\mathrm{\dots <br><br>}<br><br>,}<br><br>m_1\rangle <br><br><br><br>$

is the probability that cell (i,j) is in a state l over a time frame representing a group of samples at corresponding times m₁ through m_t, where pm,|l_i,j is the probability of obtaining the sample m_t, given that the cell (i,j) is in the state l_i,j, where pl_i,j|m_{t-1, . . . ,} m₀ is the probability of observing the cell (i,j) in state l_i,j over the time frame representing a group of samples from m_t-1 until m_o, and where p (m_t) is a normalization factor or the probability of obtaining the sample at time m_t.

20. The method according to claim 18 wherein the probability levels of the radar occupancy grid are determined in accordance with the following equation:

$p<br><br>\text{〈<br><br>}<br><br>r_{i,j}<br><br><br><br>m_{t,\mathrm{\dots <br><br>}<br><br>,}<br><br>m_1\rangle <br><br>=\frac{p<br><br>\langle <br><br>m_t<br><br><br><br>r_{i,j}\rangle <br><br><br><br>p<br><br>\langle <br><br>r_{i,j}<br><br><br><br>m_{t-1,\mathrm{\dots <br><br>}<br><br>,}<br><br>m_0\rangle <br><br>}{p<br><br>\left(m_t\right)},\mathrm{where}<br><br><br><br>p<br><br>\langle <br><br>r_{i,j}<br><br><br><br>m_{t,\mathrm{\dots <br><br>}<br><br>,}<br><br>m_1\rangle <br><br>$

is the probability that cell (i,j) is in a state r over a time frame representing a group of samples at corresponding times m₁ through m_t, where pm_t|r_i,j is the probability of obtaining the sample m_t, given that the cell (i,j) is in the state r_i,j, where pr_i,j|m_{t-1, . . . ,} m₀ is the probability of observing the cell (i,j) in state r_i,j over the time frame representing a group of samples from m_t-1 until m_o, and where p (m_t) is a normalization factor or the probability of obtaining the sample at time m_t.

21. The method according to claim 18 wherein the evaluating comprises applying a third data processing technique in accordance with Bayes law with respect to the laser occupancy grid and the radar occupancy grid to form a composite evidence grid for an overlapping region defined by an intersection of the first zone and the second zone.

22. The method according to claim 21 wherein the composite occupancy grid or precursor to the composite evidence grid is determined by the following equation:

$p<br><br>\text{〈<br><br>}<br><br>c_{i,j}<br><br><br><br>l_{i,j,}<br><br>r_{i,j}\rangle <br><br>=\frac{p<br><br>\left(c_{i,j}\right)<br><br>p<br><br>\langle <br><br>l_{i,j,}<br><br>r_{i,j}<br><br><br><br>c_{i,j}\rangle <br><br>}{p<br><br>\left(l_{i,j,}<br><br>r_{i,j}\right)},\mathrm{where}<br><br><br><br>p<br><br>\langle <br><br>c_{i,j}<br><br><br><br>l_{i,j,}<br><br>r_{i,j}\rangle <br><br>$

is the probability that cell (i,j) of the composite occupancy grid is in a state c (e.g., occupied empty or unknown) given the laser probability grid is in state l_i,j and given the radar occupancy grid is in state r_i,j, where p(c_i,j) is the probability of obtaining that cell (i,j) is in the state c_i,j, where pl_i,j,r_i,j|c_i,j is the probability of observing the cell (i,j) of the laser occupancy grid in state l and of the radar probability grid in state r, given that the composite occupancy grid is in state c, and where p(l_i,j,r_i,j) is the probability of obtaining the entries in cell (i,j) of the laser occupancy grid in state l and of the radar occupancy grid in state r.

23. The method according to claim 17 further comprising transforming probability levels of the cells of the laser occupancy grid for a non-overlapping region within the first zone to form an addendum evidence grid;

and combining the addendum evidence grid to the composite evidence grid to form an augmented evidence grid.