WO2014108889A1

WO2014108889A1 - A method for mitigating rain clutter interference in millimeter-wave radar detection

Info

Publication number: WO2014108889A1
Application number: PCT/IL2013/050039
Authority: WO
Inventors: Ehud Fishler
Original assignee: MANTISSA Ltd
Current assignee: MANTISSA Ltd
Priority date: 2013-01-14
Filing date: 2013-01-14
Publication date: 2014-07-17
Anticipated expiration: 2015-07-14

Abstract

Method for object detection in rain environment with W-band radar detection system. Radar data of region of interest is obtained, and range-velocity map generated. Initial threshold values are calculated in the range-velocity map. A rain condition is established in the region of interest based on amount of cells in the range velocity map exceeding initial threshold values. If rain condition is present, rain cells and object cells are designated in the range-velocity map, the designated rain cells forming a rain cell pattern. Rain cell threshold values are calculated in rain cell pattern. A first object cluster is detected within the rain cell cluster, a second object cluster is detected outside the rain cell pattern in the vicinity of the first object cluster, and the first object cluster is merged with the second objet cluster to form a unified object pattern. Required information associated with the object of interest is obtained.

Description

A METHOD FOR MITIGATING RAIN CLUTTER INTERFERENCE IN MILLIMETER-WAVE RADAR DETECTION

FIELD OF THE DISCLOSED TECHNIQUE

The disclosed technique generally relates to millimeter-wave radar, and more particularly, to mitigating rain clutter interference in millimeter-wave radar detection.

BACKGROUND OF THE DISCLOSED TECHNIQUE

Radar transceivers implemented in a single chip using radio frequency integrated circuit (RFIC) technology have been developed for various automotive applications, such as adaptive cruise control systems, collision avoidance systems, collision warning systems, lane change assistance systems, and parking assistance systems. W-band millimeter- wave radar systems, which transmit electromagnetic waves in the frequency range of 75GHz-1 1 0GHz, offer better range-resolution characteristics than lower frequency microwave radar, and can penetrate fog, smoke and other atmospheric obscurants significantly better than infrared sensors. A low attenuation atmospheric window of the W-band spectral range increases the operational range for a given transmitted power. Due to these features and the emerging low cost and compact device technologies, W-band millimeter-wave radar systems are employed in a wide range of commercial, military and scientific applications, including robotic vision, speed and range measurements for industrial uses, air turbulence measurements, and security applications.

One drawback of millimeter-wave radar systems is that the radio waves in this frequency range are reflected by raindrops, since the wavelength of the radio waves is of the same order of magnitude as the raindrop dimensions. This is likely to increase the false alarm rate (FAR) of an outdoor radar detection system during rainy conditions, unless the radar detection system differentiates between radar reflection from raindrops and radar reflection from an object of interest. The degree of attenuation in the radar signal resulting from rain is directly correlated to the severity of the rainfall. Therefore, the effect of rain clutter in a millimeter-wave radar detection system is more prominent during heavier rain conditions. Reference is now made to Figure 1 , which is a schematic illustration of a graph, generally referenced 1 0, of the atmospheric attenuation of radio waves as a function of frequency, for different rain conditions. The vertical axis of graph 1 0 represents attenuation in units of dB/km. The horizontal axis of graph 1 0 represents frequency ranging between 3GHz-1 00GHz and displayed on a logarithmic scale. Each graph line, of the seven graph lines depicted, is associated with a different rain condition. The rain conditions vary in severity between a drizzle (least severe rain condition), a light rain, a medium rain, a heavy rain, and a tropical downpour (most severe rain condition). Considering the W-band frequency range of 75GHz-1 00GHz, it is observed that the atmospheric attenuation ranges between approximately 0.2dB/km during drizzle conditions to nearly 1 00dB/km during tropical downpour conditions. Yet a more substantial effect of rainfall over the atmospheric attenuation of millimeter-wave radar is that the transmitted radio waves are reflected by the rain. It is thus apparent that the wave reflections from raindrops must be taken into account in order to provide reliable object detection with millimeter-wave radar systems.

The substantial signal intensity detected from raindrop reflections can significantly increase the system false alarm rate (i.e., number of false positive detections) if the rain is not identified and accounted for. Reference is now made to Figure 2, which is a schematic illustration of a graph, generally referenced 20, of the receiver radar signal intensity in the presence of rain, as a function of range. The horizontal axis of graph 20 represents range in units of meters (m), and the vertical axis of graph 20 represents the signal intensity level at the radar receiver in units of decibels (dB). It is apparent from the intensity of signal 22 that signal 22 was reflected from an object in the scene. If threshold value 21 is derived by a common processing method that provides a threshold value mostly determined by the system thermal noise and regular reflections from the ground surface, several reflections from the raindrops will exceed the threshold value and create false alarms.

The rain clutter interference problem has been addressed in the past, in attempting to obtain robust detection performance of outdoor radar systems. For instance, U.S. Patent No. 6,1 27,965 to McDade et al, entitled "Method and apparatus for rejecting rain clutter in a radar system", discloses a method and apparatus for detecting the presence of objects in a vehicle operator's blind spots. The side facing Frequency Modulated Continuous Wave (FMCW) radar detects targets even when operated in rainy weather conditions and will not generate false warnings due to rain clutter. The radar system indicates that a target is detected if and only if any part of the target is within the detection zone. By rejecting targets that are closer than the minimum range to the antenna, false alarms due to rain clutter are radically reduced.

U.S. Patent No. 6,937,1 85 to Collazo et al, entitled "Rain versus target discrimination for Doppler radars", discloses a method for better discriminating targets in rain conditions. A range-frequency map is calculated from the radar data, and cells associated with signals having a higher spectral power than noise, corresponding to reflections from raindrops, are identified on the map. The signals of rain-cells are replaced by thermal noise signals, thus reducing the false alarms associated with the rain signal. SUMMARY OF THE DISCLOSED TECHNIQUE

In accordance with one aspect of the disclosed technique, there is thus provided a method for object detection in a rain environment with a W-band radar detection system. The method includes the procedures of obtaining radar data of a region of interest with the W-band radar detection system, and generating a range-velocity map from the radar data. The method further includes the procedures of calculating initial threshold values in the range-velocity map using a constant false alarm rate processing technique, and establishing if a rain condition is present in the region of interest based on the amount of cells in the range velocity map exceeding the initial threshold values. If a rain condition is present, the method further includes the procedures of designating rain cells and object cells in the range velocity map, the designated rain cells forming a rain cell pattern, calculating rain cell threshold values in the rain cell pattern using the constant false alarm rate processing technique, detecting a first object cluster of an object of interest within the rain cell pattern, detecting a second object cluster of the object of interest outside the rain cell pattern in the vicinity of the first object cluster, and merging the first object cluster with the second objet cluster to form a unified object pattern. The method further includes the procedure of obtaining required information associated with the object of interest. In accordance with a further aspect of the disclosed technique, there is thus provided a W-band radar detection system for object detection in a rain environment. The W-band radar detection system includes a radar transceiver and a processor. The radar transceiver obtains radar data of a region of interest. The processor generates a range-velocity map from the radar data, calculates initial threshold values in the range-velocity map using a constant false alarm rate processing technique, and establishes if a rain condition is present in the region of interest based on the amount of cells in the range velocity map exceeding the initial threshold values. If a rain condition is present, the processor designates rain cells and object cells in the range velocity map, the designated rain cells forming a rain cell pattern, calculates rain cell threshold values in the rain cell pattern using the constant false alarm rate processing technique, detects a first object cluster of an object of interest within the rain cell pattern, detects a second object cluster of the object of interest outside the rain cell pattern in the vicinity of the first object cluster, merges the first object cluster with the second objet cluster to form a unified object pattern, and obtains required information associated with the object of interest. BRIEF DESCRIPTION OF THE DRAWINGS

The disclosed technique will be understood and appreciated more fully from the following detailed description taken in conjunction with the drawings in which:

Figure 1 is a schematic illustration of a graph of the atmospheric attenuation of radio waves as a function of frequency, for different rain conditions;

Figure 2 is a schematic illustration of a graph of the receiver radar signal intensity in the presence of rain as a function of range;

Figure 3 is a block diagram of a radar detection system, constructed and operative in accordance with an embodiment of the disclosed technique;

Figure 4 is a schematic illustration of two graphs, the first graph depicting the transmitted frequency waveform and the received frequency waveform of the radar transceiver of Figure 3, the second graph depicting the baseband frequency waveform of the radar transceiver of Figure 3;

Figure 5 is a flow diagram of a method for object detection in a rain environment with a millimeter-wave radar detection system, operative in accordance with an embodiment of the disclosed technique;

Figure 6 is a schematic illustration of a range-velocity map showing an object cell pattern of an object of interest detected by the radar detection system of Figure 3, in accordance with an embodiment of the disclosed technique;

Figure 7 is a schematic illustration of a range-velocity map showing a rain cell pattern of the reflection from rain as detected by the radar detection system of Figure 3, in accordance with an embodiment of the disclosed technique; and

Figure 8 is a schematic illustration of a range-velocity map showing the rain cell pattern of Figure 7 and an object cell pattern of an object of interest having a first cluster within the rain cell pattern and a second cluster outside the rain cell pattern, in accordance with an embodiment of the disclosed technique.

DETAILED DESCRIPTION OF THE EMBODIMENTS

The disclosed technique provides a method for obtaining data associated with an object of interest under rainy conditions using a millimeter-wave radar detection system, while mitigating the effect of rain clutter in the reflected radar signal. The disclosed technique is applied to a W-band, repetitive linearly frequency modulated continuous-wave (FMCW) type radar. FMCW radar facilitates processing techniques which derive simultaneously the range and velocity of the detected objects, displayed as cell patterns in two-dimensional cell arrays known as range- velocity maps. A cell pattern associated with rain is used to apply an object detection scheme that is minimally affected by the radar clutter resulting from the rain.

Reference is now made to Figure 3, which is a block diagram of a radar system, generally referenced 200, constructed and operative in accordance with an embodiment of the disclosed technique. Radar system 200 includes a radar transceiver 220 (enclosed in a dashed line), an analog to digital converter (ADC) 205 and a signal processor and radar controller 207. Radar transceiver 220 includes a radar receiver antenna

201 , a mixer 202, a baseband filter and amplifier 203, a radar transmitter antenna 21 0, a frequency modulator 21 2, and a radio frequency (RF) source 21 4. Mixer 202 is coupled with receiver antenna 201 , with RF source 21 4, and with baseband filter and amplifier 203. Baseband filter and amplifier 203 is further coupled with ADC 205. ADC 205 is further coupled with signal processor and radar controller 207. Signal processor and radar controller 207 is further coupled with frequency modulator 21 2. Frequency modulator 21 2 is further coupled with RF source 21 4. RF source 21 4 is further coupled with transmitter antenna 21 0. Radar transceiver 220 is similar to common such radar transceivers known in the art.

RF source 21 4 generates radio waves (RF signals) in the W-band frequency range (75-1 00GHz). The RF signals are modulated by frequency modulator 21 2 and transmitted by transmitter antenna 21 0 toward a region of interest. Radar receiver antenna 200 receives the radio waves reflected back from objects located in the region of interest.

Reference is now made to Figure 4, which is a schematic illustration of two graphs, the first graph referenced 31 0, depicting the transmitted frequency waveform and the received frequency waveform of the radar transceiver (220) of Figure 3, the second graph, referenced 320, depicting the waveform of the baseband frequency of the radar transceiver (220) of Figure 3. Graph 31 0 depicts in bold lines a frequency modulated (FM) transmitted radar waveform, generally referenced 31 9, and depicts in dotted lines the associated received radar waveform, generally referenced 309. The received radar waveform 309 is a replication of the transmitted radar waveform 31 9 delayed by the travel time of the transmitted wave to the object and back from the object to receiver antenna 201 . Accordingly, received radar waveform first portion 31 1 is the reflected replication of transmitted radar waveform first portion 301 . Received radar waveform second portion 31 2 is the reflected replication of transmitted radar waveform second portion 302. Similarly, received radar waveform Nth portion 31 3 is the reflected replication of transmitted radar waveform Nth portion 303. Thus, the time delay between each transmitted radar waveform portion and the associated received radar waveform potion equals the time required for the radar wave to propagate from the transmitter antenna 21 0 to the object and back to receiver antenna 201 .

Center frequency 31 5 is the non-modulated frequency of the radar transmitter antenna 200. Preferably, a center frequency of approximately 77GHz is used in accordance with the disclosed technique. The linearly shaped transmitted radar waveform 31 9 implies that the radar transmission frequency changes linearly around center frequency 31 5. The modulated frequency range determined by FM modulator 21 2 and time period of the linear portion of the frequency sweep are selected to include the maximal ranges and Doppler frequencies of an object that will be detected by the radar. It should be noted that other forms of frequency modulation (FM) applied to the transmitted RF signal are also within the scope of the disclosed technique. For instance, waveforms 31 9 and 309 may be triangular or of any other functional shape. It is further noted that radar receiver antenna 201 and radar transmitter antenna 21 0 may be embodied by a single antenna in conjunction with a circulator or a coupler array (or other coupling device), which isolates between the transmitted signal and the received signal.

A replication of the transmitted radar waveform and the received radar waveform enters mixer 202 (Figure 3), which outputs the frequency difference between transmitted radar waveform 31 9 and received radar waveform 309. The frequency difference, known as the baseband frequency, is substantially lower than the center frequency 31 5. Graph 320 depicts the baseband frequency waveform, generally referenced 329, associated with transmitted radar waveform 31 9 and received radar waveform 309 of graph 31 0. Frequency 321 of baseband frequency waveform 329 is the frequency difference between transmitted waveform first portion 301 and received waveform first portion 31 1 . Frequency 322 of baseband frequency waveform 329 is the frequency difference between transmitted waveform second portion 302 and received waveform second portion 31 2. Frequency 323 of baseband frequency waveform 329 is the frequency difference between transmitted waveform Nth portion 303 and received waveform Nth portion 31 3.

Referring back to Figure 3, the frequency output of mixer 202 (i.e., baseband frequency waveform 329) enters baseband filter and amplifier 203, which includes a band-pass filter adaptable to the baseband frequency range. ADC 205 samples the voltage level output of baseband filter and amplifier 203 at a constant rate, and these voltage samples are received by signal processor and radar controller 207. If "M" represents the number of voltage samples per transmitted and received radio waves during a certain period, and if "N" represents the number of transmitted radio waves during that period, then the output of the radar transceiver 220 (digitized by ADC 205 and arranged by signal processor and radar controller 207) is a two-dimensional digital data array of M x N dimensions. Further radar signal processing is derived by generating a range-velocity map from the two-dimensional digital data array. The range-velocity map is commonly instrumental in analyzing detected object characteristics. The horizontal axis of the map represents velocity (range rate) and the vertical axis represents range. The map is made up of an array of cells, hence the horizontal dimension of a cell represents the velocity (range rate) resolution unit, and the vertical dimension of a cell represents the range resolution unit. The features on the range-velocity map are further analyzed to derive data associated with the detected object of interest, such as the type of object, as will be elaborated upon below.

Reference is now made to Figure 5, which is a flow diagram of a method for object detection in a rain environment with a millimeter-wave radar detection system, operative in accordance with an embodiment of the disclosed technique. The method begins, in procedure 700, with obtaining radar data of a region of interest with a W-band radar detection system. Referring to Figure 3, radar transceiver 220 transmits W-band RF signals toward a region of interest, and receives the reflected RF signals. Radar detection system 200 extracts radar data associated with the transmitted and received RF signals, including digital samples of baseband filter and gain amplifier 203 or receiver 201 output frequency, as discussed hereinabove.

The method follows, in procedure 702, with generating a range- velocity map from the radar data. Referring to Figure 3, a range-velocity map is generated by applying a two dimensional fast Fourier transform (FFT) to a digital data array that is generated by ADC 205. An FFT is applied to the M samples of each of the array rows associated with each of the N frequency sweeps of the radar, followed by an FFT that is applied to each of the N columns of the array. This two level FFT operation is commonly used with FMCW radar systems to process the data into range- velocity information of an object. The complex number output elements of the two-dimensional FFT are converted into range-velocity absolute values by the square root of the sum of the real element squared and the imaginary element squared (absolute value of a complex number). It is possible to use alternative methods for unfolding the range-velocity data, such as via direct algebraic calculation from two selected sweep times of the receiver output.

Reference is now made to Figure 6, which is a schematic illustration of a range-velocity map, generally referenced 550, showing an object cell pattern, referenced 500, of an object of interest detected by the radar transceiver of Figure 3 during clear weather conditions, in accordance with an embodiment of the disclosed technique. Range- velocity map 550 represents an exemplary range-velocity map generated by radar transceiver 220 during clear weather conditions (i.e., not during rainy conditions). According to map 550, the range of detected object 500 is approximately 50m and the velocity of the object varies between approximately -2m/sec to +3m/sec, which is a characteristic Doppler spread of a walking person. It is appreciated that the detected object 500 is clearly visible on map 550.

The method follows, in procedure 704, with calculating initial threshold values in the range-velocity map. This calculation is carried out by deriving range and velocity signal intensity averages of the neighboring cells of each particular cell of the range-velocity map, and then establishing for each cell an adaptive and localized threshold value by using what is known as a Constant False Alarm Rate (CFAR) detection principle. A CFAR detection principle is a data-dependent processing technique designed to identify objects in an environment with varying background noise, by calculating the average noise level at a plurality of neighboring map cells surrounding each of the map cells and setting the detection threshold at a value of predetermined ratio to the calculated average noise. Consequently, the false alarm rate is maintained substantially fixed, while providing a high detection probability for a signal reflected from an object.

In procedure 706, it is established whether or not a rain condition is present in the region of interest, based on the amount of cells in the range-velocity map exceeding the initial threshold values. Referring to Figure 6, if the number of cells in range-velocity map 550 which have signal intensity values that exceed the respective initial threshold value is greater than a predetermined amount (e.g., or if the percentage of such cells relative to the other cells in the map exceeds a predetermined ratio value), then it is concluded that a rain condition exists.

If it is not raining (i.e., if a rain condition is determined to not exist in procedure 706), then the method follows, in procedure 708, with detecting an object of interest, using standard radar detection techniques for object detection. Subsequently, in procedure 71 0, any required information associated with the object of interest is obtained, such as the range, the velocity and the angle of the object. The range and velocity data can be obtained directly from the range-velocity map. The angle (i.e., the spatial angle of the object relative to the radar system) can be obtained in several different ways. For example, a low resolution angle estimation approach based on the specific beam direction may be used. Alternatively, the received amplitudes in neighboring/partially overlapping antenna beams may be compared, or the comparison can be done over the phase of the received signal.

If it is raining (i.e., if a rain condition is determined to exist in procedure 706), then the method follows with procedure 720. In procedure 720, rain cells and object cells are designated in the range- velocity map. From all the cells in the range-velocity map having a signal intensity value above the respective initial threshold value, if the cell is adjacent or contiguous with other such cells on the map, then it is designated as a "rain cell", whereas if the cell location is discrete or separate from other such cells on the map, then it is designated as an "object cell". The collection of designated rain cells forms a contiguous "rain cell pattern". Reference is now made to Figure 7, which is a schematic illustration of a range-velocity map, referenced 650, showing a rain cell pattern, referenced 600, of the reflection from rain as detected by the radar detection system of Figure 3, in accordance with an embodiment of the disclosed technique. The designated rain-cells are depicted in Figure 7 as the darker pixels, arranged consecutively in the form of rain- cell pattern 600. Rain cell pattern 600 is characterized by a "snake like" shape, which is a pattern typical of windy or gusty weather conditions. The range characteristic of rain cell pattern 600 extends from the minimum range to the maximum range of the radar, since rain drops are present in the entire radar depth of field. The velocity of rain cell pattern 600 varies in a substantially sinusoidal manner between approximately -2m/sec and approximately +2m/sec. The changing winds along the depth of field results in the rain cell pattern generally defining a smooth shape, such as the sinusoidal shape of rain cell pattern 600. It is appreciated that alternative rain cell patterns may also be obtained, and the rain cell pattern does not necessarily extend from the minimum to maximum range (if for example, rain is not present everywhere in the radar depth of field). Furthermore, other exemplary rain cell patterns may correspond to different velocities (i.e., beyond the range of -2m/sec to +2m/sec). The cells in map 650 which exceed the respective initial threshold value but which do not fall within rain cell pattern 650 are considered potential object cells.

In procedure 722, rain cell threshold values are calculated in the range-velocity map, using a CFAR processing technique. An updated threshold is calculated for the designated rain cells in order to detect objects located within the rain cell pattern itself. The updated threshold values are calculated from the average noise level of neighboring rain- cells of each cell in the rain-cell pattern, similar to the process described hereinabove with reference to procedure 704. The rain cell threshold values are generally higher than the initial calculated threshold values, and are generally set at a few decibels higher than the calculated average background noise.

The rain is generally present everywhere in the region of interest, including in the vicinity of the object of interest. In addition, the object typically extends over several cells in the range-velocity map. Therefore, a section of an object (or an entire object) may appear within the rain-cell pattern portion of the range-velocity map. Accordingly, it is necessary to identify any cells belonging to potential objects of interest within the previously designated rain cell pattern. In procedure 724, a first object cluster is detected within the designated rain cell pattern. The signal intensity values within the rain cell pattern are analyzed, and a cell that exceeds the respective rain cell threshold value (established in procedure 722) is identified as being associated with an object cell. The signal intensity associated with an object cell will generally be sufficiently high to exceed even the (higher) updated rain cell threshold value, and so an object cell will likely be identified even within a rain cell pattern.

In procedure 726, a second object cluster is detected outside of the designated rain cell pattern. The signal intensity values outside the rain cell pattern are analyzed, and a cell that exceeds the initial respective threshold value (established in procedure 704) is identified as being associated with an object cell. Reference is now made to Figure 8, which is a schematic illustration of a range-velocity map, referenced 800, showing the rain-cell pattern 801 which is equivalent to rain-cell pattern 600 of Figure 7 and a cell pattern of an object of interest 81 0 having a first cluster within the rain-cell pattern and a second cluster outside the rain-cell pattern, in accordance with an embodiment of the disclosed technique. Object 81 0 consists of a first cluster 81 2 (depicted by a dashed line) and a second cluster 81 1 (depicted by a solid line). First object cluster 81 2 is located and detected within rain-cell pattern 801 and second object cluster 81 1 is located and detected outside rain-cell pattern 801 .

In procedure 728, the first object of interest cluster (detected in procedure 724) and the second object of interest cluster (detected in procedure 726) are merged to form a unified object pattern, using an object cell merging algorithm. For example, edge cells belonging to one cluster of the object of interest are repositioned adjacent to the corresponding edge cells of another cluster of the object of interest. Referring to Figure 8, first object cluster 81 2 and second object cluster 81 1 are merged, to provide the full object cell pattern of the entire object of interest 81 0. It is further noted that the merging algorithm eliminates "ghost targets" (i.e., when the same object cell is identified twice) and ensures an accurate calculation of the unified object parameters (range, velocity and angle). For example, if there exists a single object in the scene, then the same object cell could potentially be detected twice in neighboring cells: once during the initial application of the CFAR processing of the range-velocity map (procedure 704) at a first cell coordinates (e.g., X,Y), and then a second time during the application of the CFAR processing of the rain cell pattern (procedure 722) at a slightly displaced location (e.g., X+1 ,Y). In such a case, the same object cell would be mistakenly considered as belonging to two separate objects with two distinct sets of object parameters. Accordingly, the merging algorithm identifies such a scenario and outputs the associated object cell at a single location (e.g., at cell coordinates X+0.5, Y), ensuring correct detection of a single unified object.

The method resumes with procedure 71 0, in which required information associated with the object of interest is obtained (as discussed previously).

It will be appreciated by persons skilled in the art that the disclosed technique is not limited to what has been particularly shown and described hereinabove.

Claims

CLAIMS A method for object detection in a rain environment with a W-band radar detection system, the method comprising the procedures of: obtaining radar data of a region of interest with said W-band radar detection system;

generating a range-velocity map from said radar data;

calculating initial threshold values in said range-velocity map using a constant false alarm rate (CFAR) processing technique;

establishing if a rain condition is present in said region of interest, based on the amount of cells in said range-velocity map exceeding said initial threshold values;

if a rain condition is present:

designating rain cells and object cells in said range-velocity map, said designated rain cells forming a rain cell pattern;

calculating rain cell threshold values in said rain cell pattern using said CFAR processing technique;

detecting a first object cluster of an object of interest within said rain-cell pattern;

detecting a second object cluster of said object of interest outside said rain cell pattern in the vicinity of said first object cluster; merging said first object cluster with said second object cluster to form a unified object pattern;

and

obtaining required information associated with said object of interest.

2. The method according to claim 1 , wherein said range-velocity map is obtained by applying a first fast Fourier transform (FFT) to each of the rows of a two-dimensional sampled data matrix associated with said radar data, applying a second FFT to the columns of said data matrix, and calculating absolute values of the complex number elements of said data matrix.

3. The method according claim 1 , wherein said initial threshold values are determined as a function of the average thermal noise level of the radar signal of said W-band radar detection system.

4. The method according to claim 1 , wherein said procedure of establishing if a rain condition is present comprises counting the number of cells which exceed the respective said initial threshold value, and verifying that said number of cells is above a predetermined threshold number.

5. The method according to claim 1 , wherein said rain cell threshold values are determined as a function of the average signal level of neighbouring rain cells within said rain cell pattern.

6. The method according to claim 1 , wherein said required information associated with said object of interest comprises at least one of:

the range of said object of interest;

the velocity of said object of interest; and

the angle of said object of interest.

7. The method according to claim 1 , wherein said rain cells are substantially contiguous in said range-velocity map, and said object cells are substantially discrete in said range-velocity map.

8. The method according to claim 1 , wherein said rain cell pattern defines a substantially smooth shape.

9. The method according to claim 1 , wherein said rain cell pattern defines a snake-like shape.

0. The method according to claim 1 , wherein said rain condition comprises a weather condition selected from the list consisting of: drizzle;

light rain;

heavy rain;

downpour;

thunderstorm;

snow;

hail;

sleet; and

any combination of the above. 1 . The method according to claim 1 , wherein said W-band radar detection system operates at a frequency of approximately 77 GHz. 2. A W-band radar detection system for object detection in a rain environment, the system comprising:

a radar transceiver, operative to obtain radar data of a region of interest; and

a processor, coupled with said radar transceiver, operative to generate a range-velocity map from said radar data, to calculate initial threshold values in said range-velocity map using a constant false alarm rate (CFAR) processing technique, and to establish if a rain condition is present in said region of interest, based on the amount of cells in said range-velocity map exceeding said initial threshold values,

wherein if a rain condition is present, said processor designates rain cells and object cells in said range-velocity map, said designated rain cells forming a rain cell pattern, calculates rain cell threshold values in said rain cell pattern using said CFAR processing technique, detects a first object cluster of an object of interest within said rain-cell pattern, detects a second object cluster of said object of interest outside said rain-cell pattern in the vicinity of said first object cluster, merges said first object cluster with said second object cluster to form a unified object pattern, and obtains required information associated with said object of interest. 3. The radar detection system according to claim 1 2, wherein said radar transceiver comprises:

a radio frequency (RF) source, operative to generate RF signals in the W-band frequency range;

a frequency modulator, operative to modulate said RF signals; a radar transmitter antenna, operative to transmit the modulated RF signals toward said region of interest; and a radar receiver antenna, operative to receive reflected waves of said transmitted signals. 4. The radar detection system according to claim 1 3, wherein said radar transmitter antenna and said radar receiver antenna comprises a single antenna and a coupling device, operative to isolate between said transmitted signals and said received signals. 5. The radar detection system according to claim 1 3, wherein said radar transceiver further comprises a mixer, operative to provide a baseband frequency waveform comprising the frequency difference between the waveform of said transmitted signals and the waveform of said received signals. 6. The radar detection system according to claim 1 5, wherein said radar transceiver further comprises a baseband filter and amplifier, operative to filter and amplify said baseband frequency waveform. 7. The radar detection system according to claim 1 6, further comprising an analog to digital converter (ADC), coupled with said radar transceiver and said processor, said ADC operative to sample said filtered and amplified baseband frequency.

18. The radar detection system according to claim 12, operative at a frequency of approximately 77 GHz.