Granted Patent B2

US 12,688,700 · App. 18/567,620 · Granted Jul 21, 2026

Inventors: Mickael Mimoun (Bobigny, FR); Yasser Almehio (Bobigny, FR)

Assignee: Valeo Vision

G06V20/58B60Q1/0023B60Q1/04G01B11/25G01B11/2518G06V10/145

Abstract

Claims (38)

1 . A method of detecting an object in a road surface, comprising:

an autonomous vehicle system;

a luminous device having a plurality of light pixels;

an image sensor;

a processing unit applying a machine learning process that controls:

wherein the luminous device projects a light pattern on the road surface;

wherein the image sensor acquires an image of the projected light pattern;

wherein the image sensor detects a shadow in the image being acquired; and

wherein the processing unit uses a portion of features from the shadow to obtain information about object features being detected in order to control results of a modified light pattern being projected on the road surface.

2 . The method according to claim 1 , wherein the light pattern is a high beam pattern.

3 . The method according to claim 1 , wherein the light pattern is a uniform pattern.

4 . The method according to claim 1 , wherein the light pattern is projected by a headlamp.

5 . The method according to claim 1 , wherein the light pattern is projected by a reverse light.

6 . The method according to claim 1 , further comprising modifying the light pattern when an object is detected.

7 . The method according to claim 6 , wherein the modified light pattern includes light stripes.

8 . The method according to claim 7 , wherein the light stripes are horizontal, vertical or diagonal light stripes.

9 . The method according to claim 6 , wherein the modified light pattern has a higher luminous intensity than the original light pattern in the zone where the object is detected.

10 . The method according to claim 1 , wherein the features of the shadow contain the position, the width and/or the height of the object.

11 . The method according to claim 1 , further comprising

providing a lighting device with a database of debris objects, wherein the database contains the debris objects about different sizes, materials, shapes, orientations and additional shadows;

wherein utilizing a portion of features of the shadow to obtain information about the object features is carried out by the machine learning process; and

the machine learning process includes a pre-processing of each image, which includes an image equalization to enhance a contrast between a lighted surface and the shadow created thereby.

12 . A method for managing an autonomous vehicle, comprising:

a luminous device, a plurality of light pixels, an image sensor;

a processing unit with a machine learning process that controls the processing unit to perform the following:

performing detection of an object through the image sensor combined with projecting a light pattern through the luminous device on a road surface;

acquiring an image of the light pattern being projected through the image sensor;

detecting a shadow in the image being acquired; and

utilizing a portion of features from the shadow to obtain information about object features;

utilizing the object features through the machine learning process to decide a suitable vehicle maneuver;

checking through the machine learning process if the suitable vehicle maneuver can be performed in secure conditions; and

performing the suitable vehicle maneuver through the processing unit controlling an autonomous vehicle.

13 . An automotive lighting device of an autonomous vehicle system comprising:

a plurality of solid-state light sources; an image sensor;

a processing unit having a machine learning process;

wherein the plurality of solid-state light sources are adapted to project a light pattern on a road surface controlled by the processing unit;

wherein the image sensor is controlled by the processing unit to acquire an image of the light pattern being projected; and

the processing unit is controlled to detect a shadow in the image being acquired and and uses a portion of features from the shadow to obtain information about object features being detected by the image sensor.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (23)