Image processing system, image processing method, and storage medium each for obtaining pixels of object using neural network
A first decoder obtains a region including an object targeted for recognition based on a feature map obtained by performing convolutional processing on a processing target image. Next, the first decoder obtains, from the feature map, a partial feature map of a portion corresponding to the obtained region including the object targeted for recognition. Then, a second decoder obtains pixels corresponding to the object targeted for recognition based on the partial feature map. This reduces the amount of calculation required for decoder units included in a neural network.
1. An image processing apparatus comprising:
at least one memory that stores instructions that include a neural network; and
at least one processor that executes the instructions to perform:
convolutional processing on a processing target image by an encoder included in the neural network to obtain a feature map;
first deconvolutional processing on the obtained feature map by a first decoder included in the neural network to obtain a region including an object targeted for recognition included in the processing target image;
obtaining, from the obtained feature map, a partial feature map of a portion corresponding to the obtained region; and
second deconvolutional processing on the obtained partial feature map by a second decoder included in the neural network to obtain pixels corresponding to the object targeted for recognition included in the processing target image.
2. The image processing apparatus according to claim 1 , wherein the obtained partial feature map is a partial feature map of a portion corresponding to a region excluding a region having a predetermined number or less of pixels from among the region including the object targeted for recognition obtained by the first deconvolutional processing.
3. The image processing apparatus according to claim 1 , wherein the object targeted for recognition includes a handwritten character.
4. The image processing apparatus according to claim 1 , wherein the object targeted for recognition includes an impressed seal.
5. A non-transitory computer-readable storage medium storing instructions including a neural network, wherein the instructions cause a computer to perform:
convolutional processing on a processing target image by an encoder included in the neural network to obtain a feature map;
first deconvolutional processing on the obtained feature map by a first decoder included in the neural network to obtain a region including an object targeted for recognition included in the processing target image;
obtaining, from the obtained feature map, a partial feature map of a portion corresponding to the obtained region; and
second deconvolutional processing on the obtained partial feature map by a second decoder included in the neural network to obtain pixels corresponding to the object targeted for recognition included in the processing target image.
6. An image processing method comprising:
convolutional processing on a processing target image by an encoder included in a neural network to obtain a feature map;
first deconvolutional processing on the obtained feature map by a first decoder included in the neural network to obtain a region including an object targeted for recognition included in the processing target image;
obtaining, from the obtained feature map, a partial feature map of a portion corresponding to the obtained region; and
second deconvolutional processing on the obtained partial feature map by a second decoder included in the neural network to obtain pixels corresponding to the object targeted for recognition included in the processing target image.