Image analysis – Applications – Vehicle or traffic control
Reexamination Certificate
1999-08-26
2002-09-03
Patel, Jayanti K. (Department: 2721)
Image analysis
Applications
Vehicle or traffic control
C340S436000, C340S903000, C348S118000
Reexamination Certificate
active
06445809
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an environment monitoring system for picking up the front, rear or diagonal-rear view from one's own vehicle by an image pick-up means such as a video camera installed on the vehicle such as a motor car, detecting another vehicle approaching from the front, rear or diagonal-rear direction of one's own running vehicle using the image picked up and giving a warning to a driver.
2. Description of the Related Art
For example, when a driver of a vehicle running on one-side two lanes of e.g. a speed way intends to change his own vehicle lane, if he changes the lane while he misses another vehicle which catches up with his own vehicle on another lane at a higher speed than his own vehicle from the diagonal-rear direction, there is strong possibility of a serious accident. Therefore, it is desired that the driver surely notices or recognizes other vehicles in the environment.
When another following vehicle on the same lane abruptly approaches his own vehicle from the rear, if the driver of his own vehicle makes abrupt braking, there is possibility of bumping-into-the back.
When another vehicle runs forward on the same lane at a lower speed than that of his own vehicle, if a driver dozes, there is a danger of bumping-into-the back.
Particularly, in recent years, various types of vehicles inclusive of a passenger can have a highly improved performance so that the running speed or acceleration thereof has been greatly increased. Therefore, the vehicle moves very fast in lane changing or acceleration/deceleration. As a result, when the driver misses the presence of another vehicle running in the environment of his own vehicle, there has been increased strong possibility of occurrence of a dangerous state which may lead to serious accidents, or an actual serious accident.
In order to prevent the driver from missing another in the environment than his own vehicle (hereinafter referred to as “surrounding vehicle”), the driver himself must first pay attention to the environment. However, it is true that attention or recognition ability of a human being has reached the limit of ability of capable of following the improved performance of a vehicle. Actually, there is an increased tendency of accidents attributable to the missing the surrounding vehicle at the limit value.
Generally, it is known that the visual recognition ability of a human being abruptly decreases as the running speed increases. The vehicle is still desired to deal with a higher running speed in the future. Therefore, there is a serious problem of danger of missing the presence of the surrounding vehicle in e.g. lane changing.
Accordingly, in order to prevent the danger of a great accident or the actual great accident from occurring, only enhancing the attention of the driver to the presence of the surrounding vehicle is not sufficient. It is necessary to recognize the presence of the surrounding vehicle running in the environment automatically, accurately and surely and give the acquired information to the driver so that the limited recognition ability of a human being can be complemented.
Several techniques for avoiding the danger as described above have been proposed as disclosed in JP-A-6-107096 and JP-A-7-50769. Specifically, JP-A-6-107096 discloses a system for detecting an approaching vehicle and collision warning based on the detection of an optical flow. JP-A-7-50769 discloses a warning system which intends to detect an approaching vehicle from the diagonal-rear direction on the basis of detection of an optical flow and issues a warning when there is a danger of colliding with another vehicle in lane changing.
Now referring to
FIGS. 20 and 21
, an explanation will be given of the summary of a conventional environment monitoring system.
FIGS. 20A-20D
are views for explaining a change in a rear background image acquired by a video camera
1
. FIG.
20
A shows a status inclusive of one's own vehicle.
FIG. 20B
shows an image picked up by a video camera
1
at timing t in an environment of one's own vehicle.
FIG. 20C
shows an image picked up at timing t+&Dgr;t.
Now it is assumed that one's own vehicle is running straight on a flat road. The road sign and building residing in the rear of one's own vehicle in
FIG. 20A
are observed as images shown in
FIGS. 20B and 20C
at timings t and t+&Dgr;t, respectively. Connecting the corresponding points in these two images provides speed vectors as shown in FIG.
20
D. The are referred to as “optical flows”. Where a following vehicle approaches, the directions of the vectors in the optical flows in
FIG. 20D
are contrary.
These optical flows appear radially from a point called “FOE” (Focus of Expansion) in the image. The FOE, which refers to “infinite point” or “disappearing point”, corresponds to the one point opposite to the running direction of one's own vehicle on the image where one's own vehicle runs straight. In this way, the optical flows acquired when one's own vehicle runs extends radially around the FOE. The optical flows issued from the following vehicle or another vehicle running on an adjacent lane (hereinafter referred to as “particular vehicle”) include information composed of a position and relative speed of the above particular vehicle. Where the optical flows are longer and diverge from the FOE, there is strong possibility of danger.
Now referring to
FIG. 21
, a detailed explanation will be given of the optical flows. In an optical arrangement shown in
FIG. 21
, it is assumed that reference numeral
11
denotes a lens of a video camera,
12
denotes an image plane of the video camera, f denotes a distance from the lens
11
to the image plane
12
, P (X, Y, Z) denotes any point on the following vehicle, and p (x y) denotes a point corresponding to the point P on the image plane
12
.
From the ratio of similarity of a triangle,
x=f·X/Z
(1)
By transforming and time-differentiating this equation,
x′=
(&Dgr;
x/&Dgr;t·Z+x·Z′
)
/f
(2)
The component u in the x direction of the optical flow is expressed by
u=&Dgr;x/&Dgr;t
(3)
Using it, Z can be expressed by
Z=
(
f·X′−x·Z′
)
/u
(4)
Now, Z denotes a relative speed between one's own vehicle and the “particular vehicle”. Assuming that
Z′=−&agr;,
(5)
Equation (4) can be transformed into
Z=
(
f·X′+x &agr;
)
/u
(6)
Therefore, the x direction component of the optical flow (i.e. &Dgr;x/&Dgr;t=u) can be expressed by
u=
(
f·X′+x&agr;
)
/Z
(7)
This applies to Y.
Thus, from Equation (7), as Z i.e. the distance from the particular vehicle is decreased, otherwise &agr; is increased (relative speed is large), the x component of the optical flow is increased. This applies to the Y direction.
Therefore, the optical flow is longer as the distance from the “particular vehicle” is shorter or the relative speed is higher. Accordingly, a greater length of the optical flow diverging from FOE provides higher degree of danger for the particular vehicle.
The above processing is repeated for all the points on the image at timing t so that the optical flows of the entire image representative of the danger of each particular vehicle can be acquired. An alarm is sounded according to the acquired degree of danger to alert the driver of his own vehicle to the danger. This complements the limited recognition ability of a human being, and prevents the danger of a great accident or the actual occurrence thereof.
In a prior art, as shown in
FIG. 22
, in order to save the processing time for an object for which monitoring is not required, the white lines of a lane on which one's own vehicle runs on a straight road are detected so that the lane on which one's own vehicle runs is distinguished from an adjacent lane to define a monitoring region. The FOE point is acquired from the extension
Imai Teruhisa
Ishikawa Naoto
Nakajima Masato
Sasaki Kazuyuki
Armstrong Westerman & Hattori, LLP
Patel Jayanti K.
Yazaki -Corporation
LandOfFree
Environment monitoring system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Environment monitoring system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Environment monitoring system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2882344