Image processing apparatus,image processing method, and program

1. An image processing apparatus comprising:

a generation unit configured to generate a background model of a multinomial distribution from an acquired image, and generate a background image using the background model;

a moving object detection unit configured to detect a moving object in the acquired image, wherein the moving object is an area in the acquired image different from the background image; and

a reliability processing unit configured to calculate a value for a pixel of the acquired image indicating a probability of the pixel being a non-moving object using the acquired image, the background image and a temporarily paused image, wherein the temporarily paused image comprises the moving object which is paused for at least a predetermined time period,

wherein the background model includes a distribution in which a luminance value at a predetermined coordinate is associated with an appearance frequency of the luminance value, and wherein a determination unit determines the predetermined coordinate as a processable area when the distribution with highest appearance frequency among distributions in the multinomial distribution is larger than a predetermined value,

wherein the predetermined value is calculated from a number of terms of the multinomial distribution and a number of updates of the background model.

2. The image processing apparatus according to claim 1 , wherein the determination unit is configured to detect an unprocessable area in the background image, wherein the detected unprocessable area in the background image is presented to a user.

3. The image processing apparatus according to claim 1 , further comprising:

a temporary pause determination unit configured to determine whether the moving object is paused for at least the predetermined time period; and

a non-moving object detection unit configured to detect the non-moving object from the acquired image based on the value, wherein the value indicates the probability of the pixel of the acquired image being the non-moving object which is an image different from the background image.

4. The image processing apparatus according to claim 1 , wherein the determination unit is configured to detect an unprocessable area in the background image, wherein the detected unprocessable area in the background image is an area in which a distribution with highest appearance frequency among distributions in the multinomial distribution is less than a predetermined value.

5. The image processing apparatus according to claim 1 , wherein the background model is generated for each of a plurality of pixels in the acquired image.

6. The image processing apparatus according to claim 1 , wherein the generation unit generates the background model of a trinomial distribution.

7. An image processing method comprising:

in an image processing apparatus comprising one or more processors:

generating a background model of a multinomial distribution from an acquired image, and generating a background image using the background model,

wherein the background model includes a distribution in which a luminance value at a predetermined coordinate is associated with an appearance frequency of the luminance value, wherein the predetermined coordinate is determined as a processable area when the distribution with highest appearance frequency among distributions in the multinomial distribution is larger than a predetermined value, and

wherein the predetermined value is calculated from a number of terms of the multinomial distribution and a number of updates of the background model.

8. The image processing method according to claim 7 , comprising:

detecting a moving object in the acquired image, wherein the moving object is an area in the acquired image different from the background image;

calculating a value for a pixel of the acquired image indicating a probability of the pixel being a non-moving object using the acquired image, the background image and a temporarily paused image, wherein the temporarily paused image comprises the moving object which is paused for at least a predetermined time period;

determining whether the moving object is paused for at least the predetermined time period; and

detecting the non-moving object from the acquired image based on the value, wherein the value indicates the probability of the pixel of the acquired image being a non-moving object which is an image different from the background image.

9. The image processing method according to claim 8 , wherein the predetermined time period is set by a user.

10. A non-transitory computer-readable recording medium having a set of instructions recorded thereon, the set of instructions causing a computer to perform steps comprising:

generating a background model of a multinomial distribution from an acquired image, and generating a background image using the background model,

wherein the background model includes a distribution in which a luminance value at a predetermined coordinate is associated with an appearance frequency of the luminance value, wherein the predetermined coordinate is determined as a processable area when the distribution with highest appearance frequency among distributions in the multinomial distribution is larger than a predetermined value, and

wherein the predetermined value is calculated from a number of terms of the multinomial distribution and a number of updates of the background model.

11. An image processing apparatus comprising:

a generation unit configured to generate a background model of a multinomial distribution from an acquired image, and generate a background image using the background model; and

a determination unit configured to detect an unprocessable area in the background image using the background model,

wherein the background model includes a distribution in which a luminance value at a predetermined coordinate is associated with an appearance frequency of the luminance value, and wherein the determination unit determines the predetermined coordinate as a processable area when a distribution with a highest appearance frequency among distributions in the multinomial distribution is greater than a predetermined value at the predetermined coordinate, and

wherein the predetermined value is calculated from a number of terms of the multinomial distribution and a number of updates of the background model.

12. An image processing apparatus comprising:

a generation unit configured to generate a background model of a multinomial distribution from an acquired image, and generate a background image using the background model;

a determination unit configured to detect an unprocessable area in the background image using the background model;

a moving object detection unit configured to detect a moving object in the acquired image, wherein the moving object is an area in the acquired image different from the background image;

a temporary pause determination unit configured to determine whether the moving object is paused for at least a predetermined time period;

a reliability processing unit configured to calculate a value for a pixel of the acquired image using the acquired image and a temporarily paused image, wherein the temporarily paused image comprises the moving object which is paused for at least the predetermined time period, wherein the value indicates a probability of the pixel of acquired image being a non-moving object which is an image different from the background image that does not change for at least the predetermined time period; and

a non-moving object detection unit configured to detect the non-moving object from the acquired image based on the value.