Method and system for detecting and recognizing target in real-time video, storage medium, and device

A system for detecting and recognizing a target object in a real-time video, comprising a circuitry to implement:

• a first-stage neural network, configured to perform target object position detection in a frame of image of a video input to obtain a target object position range;
• and
• a second-stage neural network, configured to perform target object recognition in the frame of image according to the target object position range to obtain a recognition result,
• wherein the circuitry is configured to:
• determine whether a target object recognition result RX-1 of a previous frame of image of a current frame of image is the same as a target object recognition result RX-2 of a previous frame of image of the previous frame of image;
• control the first-stage neural network to perform target object position detection in the current frame of image to obtain a position range CX of a target object in the current frame of image when the two recognition results RX-1 and RX-2 are different;
• or determine a position range CX of a target object in the current frame of image according to a position range CX-1 of the target object in the previous frame of image without using the first-stage neural network when the two recognition results RX-1 and RX-2 are the same;
• and
• control the second-stage neural network to perform target object recognition in the current frame of image according to the position range CX of the target object in the current frame of image to obtain a target object recognition result RX of the current frame of image.

A non-transitory storage medium for storing computer readable instructions, the computer readable instructions, when executed by a processor, causing the processor to:

• determine whether a target object recognition result RX-1 of a previous frame of image of a current frame of image is the same as a target object recognition result RX-2 of a previous frame of image of the previous frame of image;
• perform target object position detection in the current frame of image by using a first-stage neural network to obtain a position range CX of a target object in the current frame of image when the two recognition results RX-1 and RX-2 are different;
• or determine a position range CX of a target object in the current frame of image according to a position range CX-1 of the target object in the previous frame of image without using a first-stage neural network when the two recognition results RX-1 and RX-2 are the same;
• and
• perform target object recognition in the current frame of image according to the position range CX of the target object in the current frame of image by using a second-stage neural network to obtain a target object recognition result RX of the current frame of image.