Surveillance system with function of automatically generating text summaries and generation method thereof
A surveillance system with a function of automatically generating text summaries and a text summary generation method are provided. The surveillance system includes a portable electronic device, an image capturing device, a computing device and a signal transmission device. The image capturing device captures an image of a surveillance area. When an intelligent application image recognition program in the computing device is read and executed, the captured image of the surveillance area is processed into the corresponding text summary. In addition, the text summary is transmitted to the portable electronic device through the signal transmission device.
1 . A surveillance system with a function of automatically generating text summaries, the surveillance system comprising:
a portable electronic device;
an image capturing device capturing an image of a surveillance area;
a computing device storing an intelligent application image recognition program, wherein when the intelligent application image recognition program in the computing device is read and executed, an automatic text summary generation method is performed, wherein the automatic text summary generation method comprises steps of:
(a) setting a range of the surveillance area and a notification condition;
(b) driving the image capturing device to capture the image of the surveillance area;
(c) judging whether the notification condition is triggered in response to the captured image;
(d) if an image segment in the captured image complies with the notification condition, extracting the image segment from the captured image;
(e) divided the image segment into a series of multiple image files;
(f) extracting a subject and an action of the subject complying with the notification condition from each of the multiple image files; and
(g) generating a segment descriptive sentence corresponding to the subject and the action in each of the multiple image files, and combining the segment descriptive sentence corresponding to the image segment into a chain-type summary in a chronological order; and
a signal transmission device in communication with each of the portable electronic device, the image capturing device and the computing device,
wherein the chain-type summary is transmitted from the computing device to the portable electronic device through the signal transmission device via the communication.
2 . The surveillance system according to claim 1 , wherein in the step (a), the notification condition contains at least one object recognition feature and at least one object motion recognition feature.
3 . The surveillance system according to claim 2 , wherein the step (f) performed by the computing device comprises steps of:
(f1) extracting the subject complying with the at least one object recognition feature according to the at least one object recognition feature; and
(f2) extracting the action complying with the at least one object action recognition feature according to the at least one object action recognition feature.
4 . The surveillance system according to claim 3 , wherein the step (f) performed by the computing device further comprises a step of adding timestamps to the multiple image files respectively and sequentially.
5 . The surveillance system according to claim 3 , wherein the step (g) performed by the computing device comprises steps of:
(g1) generating a subject descriptive sentence corresponding to the subject and a motion descriptive sentence corresponding to the motion; and
(g2) combining the subject descriptive sentence and the motion descriptive sentence in the chronological order so as to form the segment descriptive sentence.
6 . The surveillance system according to claim 1 , wherein the step (e) performed by the computing device comprises steps of:
(e1) dividing the image segment into the multiple image files; and
(e2) adding timestamps to the multiple image files respectively and sequentially.
7 . The surveillance system according to claim 1 , wherein the computing device is a home computer or a cloud server.
8 . The surveillance system according to claim 1 , wherein the intelligent application image recognition program has an image recognition model database, and the image recognition model database contains a plurality of object recognition feature data and a plurality of object action recognition feature data, wherein the plurality of object recognition feature data and the plurality of object action recognition feature data are added to the image recognition model database according to input training data from the image segment.
9 . An automatic text summary generation method for analyzing an image that is captured in a surveillance area by an image capturing device, the automatic text summary generation method comprising steps of:
(a) setting a range of the surveillance area and a notification condition;
(b) driving the image capturing device to capture the image of the surveillance area;
(c) judging whether the notification condition is triggered in response to the captured image;
(d) if an image segment in the captured image complies with the notification condition, extracting the image segment from the captured image;
(e) divided the image segment into a series of multiple image files;
(f) extracting a subject and an action of the subject complying with the notification condition from each of the multiple image files; and
(g) generating a segment descriptive sentence corresponding to the subject and the action in each of the multiple image files, and combining the segment descriptive sentence corresponding to the image segment into a chain-type summary in a chronological order; and
(h) transmitting the chain-type summary to a portable electronic device.
10 . The automatic text summary generation method according to claim 9 , wherein in the step (a), the notification condition contains at least one object recognition feature and at least one object motion recognition feature.
11 . The automatic text summary generation method according to claim 10 , wherein the step (f) comprises steps of:
(f1) extracting the subject complying with the at least one object recognition feature according to the at least one object recognition feature; and
(f2) extracting the action complying with the at least one object action recognition feature according to the at least one object action recognition feature.
12 . The automatic text summary generation method according to claim 11 , wherein the step (f) further comprises a step of adding timestamps to the multiple image files respectively and sequentially.
13 . The automatic text summary generation method according to claim 11 , wherein the step (g) comprises steps of:
(g1) generating a subject descriptive sentence corresponding to the subject and a motion descriptive sentence corresponding to the motion; and
(g2) combining the subject descriptive sentence and the motion descriptive sentence in the chronological order so as to form the segment descriptive sentence.
14 . The automatic text summary generation method according to claim 9 , wherein the step (e) comprises steps of:
(e1) dividing the image segment into the multiple image files; and
(e2) adding timestamps to the multiple image files respectively and sequentially.