Image-sensing method having a very short integration time
The invention relates to image sensors, and more particularly to matrix sensors having active pixels in CMOS technology. According to the invention, a method for imaging a scene with a very short integration time, by using a standard image sensor comprising a matrix of pixels, comprising photodiodes, and charge storage nodes, is provided. Two images of the same scene are produced under identical light conditions, one of the images corresponding to integration of charges during a first time interval with a duration T int and the other image corresponding to a second time interval with a duration T′ int longer than T int , and a difference between these two images is established, representing an image integrated during a time interval T′ int −T int . The light may be provided by a light pulse (IMP). This method may be used for observing points in a scene that lie at a well-determined distance, the brevity of the integration time allowing good precision of the observation distance.
1. A method for imaging a scene with a very short integration time by using an image sensor comprising a matrix of pixels, each pixel comprising a photodiode and means for reading the charges accumulated by the photodiode during an integration duration, the matrix furthermore comprising sequencing means for establishing start and end instants of the integration duration, the method being characterized in that two images of the same scene are produced under identical light conditions, one of the images corresponding to integration of charges during a first time interval with a duration T int and corresponding to the even rows of the matrix, the other image corresponding to a second time interval with a duration T′ int longer than T int and corresponding to the odd rows of the matrix, and a difference between these two images is established, representing an image integrated during a time interval T′ int −T int , the difference being formed by the pixel-to-pixel difference between the luminances of the pixels of an odd row and the luminances of the pixels of an even row immediately adjacent.
2. The method as claimed in claim 1 , wherein a controlled light source providing light pulses calibrated in duration and in intervals in synchronism with the operation of the sensor is used to illuminate the scene, and two images of the scene, illuminated by a light pulse, are produced, one of the images corresponding to integration of charges in the pixels of the even rows during the first time interval with a duration for T int and the other image corresponding to integration of charges in the pixels of the odd rows during the second time interval with a duration T′ int longer than T int .
3. The method as claimed in claim 2 , wherein the observed scene lies at a determined distance d i , and the synchronization between the light source and the sensor is such that the end of the light pulse reflected by the observed scene and corresponding to the image integrated during the duration T int fully or partially overlaps the portion of duration T′ int that does not coincide with the duration T int .
4. The method as claimed in claim 2 , wherein the images are produced on the basis of a train of light pulses.
5. The method as claimed in claim 3 , wherein the images are produced on the basis of a train of light pulses.
6. The method as claimed in claim 1 , wherein the images are produced on the basis of a train of light pulses.
7. The method as claimed in claim 1 , wherein the first time interval with a duration T int and the second time interval with a duration T′ int are established on the basis of asynchronous logic circuits comprising CMOS inverters.