Method and system for vehicle image repositioning using machine learning
An image readjustment system featuring a communication interface, one or more processors, and a persistent storage. The persistent storage is communicatively coupled to the one or more processors. The persistent storage comprises at least image location logic and image cropping logic. The image location logic is configured to receive and analyze an incoming image, using a machine learning analysis, to determine a location of a particular component within the incoming image and to calculate coordinates of a boundary area surrounding the particular component. The image cropping logic is configured to remove portions of the incoming image outside of the boundary area to generate a modified image to be rendered for display.
1. An image readjustment system, comprising:
one or more processors; and
a persistent storage communicatively coupled to the one or more processors, the persistent storage comprises
an image location logic to receive and analyze an incoming image, using a machine learning analysis, to determine a location of a particular component within the incoming image and to calculate coordinates of a boundary area surrounding the particular component,
an aspect ratio determination logic configured to determine an aspect ratio of the incoming image,
an image repositioning logic that is configured to use both the aspect ratio determined for the incoming image and the coordinates of the boundary area to generate a secondary boundary area including the particular component substantially centered within the secondary boundary area and featuring an aspect ratio that is (i) equal to the aspect ratio of the incoming image or (ii) substantially equivalent to the aspect ratio of the incoming image as being within three-percent of the aspect ratio of the incoming image, and
an image cropping logic to remove portions of the incoming image outside of the secondary boundary area to generate a modified image to be rendered for display.
2. The image readjustment system of claim 1 , wherein the image location logic to calculate the coordinates of the boundary area and provide at least two coordinates of the boundary area to the image cropping logic.
3. The image readjustment system of claim 1 , wherein the modified image and coordinates of the particular component are stored within a data store of the persistent storage, the data store being accessible by logic for rendering the modified image.
4. The image readjustment system of claim 1 , wherein the particular component is an image of a vehicle.
5. The image readjustment system of claim 4 , wherein the persistent storage further comprises widget server logic that is configured to receive a communication from a widget, being software code residing in a website, to obtain information pertaining to the vehicle including the modified image of the vehicle.
6. The image readjustment system of claim 1 , wherein the image repositioning logic being communicatively coupled to the image cropping logic to remove portions of the incoming image outside of the secondary boundary area including at least a portion of the boundary area.
7. The image readjustment system of claim 1 , wherein the image location logic is further configured to determine whether the incoming information includes customized information and preclude generation of the boundary area and the second boundary area to avoid cropping of the incoming image to retain the customized information, the customized information including any of dealership information, salesperson information or dealer logos.
8. A computerized method operating within a server device for analyzing images using machine learning logic, trained through supervised learning, to identify and reposition a targeted component of a first component type associated with training data provided during the supervised learning, comprising:
receiving an incoming image;
analyzing the incoming image, using the machine learning logic, to determine a location of the targeted component within the incoming image including a plurality of components of the first component type and to calculate coordinates of a boundary area surrounding the targeted component;
determining an aspect ratio of the incoming image;
determining a secondary boundary area based on the aspect ratio and the coordinates of the boundary area determined for the incoming image, wherein the targeted component is substantially centered within the secondary boundary area and the secondary boundary area featuring an aspect ratio that is (i) equal to the aspect ratio of the incoming image or (ii) substantially equivalent to the aspect ratio of the incoming image as being within three-percent of the aspect ratio of the incoming image; and
removing portions of the incoming image outside of the secondary boundary area to generate a modified image to be rendered for display.
9. The computerized method of claim 8 , wherein the calculating of the coordinates of the boundary area comprises determining at least two coordinates to form the boundary area.
10. The computerized method of claim 8 , wherein the removing of the portion of the incoming image outside of the secondary boundary area occurs after centering the targeted component within the secondary boundary area.
11. The computerized method of claim 10 , wherein the targeted component is an image of a vehicle appearing to be a key component of the image and the first component type is directed to vehicles.
12. The computerized method of claim 11 further comprising:
storing the modified image within a data store accessible by the server device;
retrieving the modified image from the data store in response to widget server logic, implemented within the server device and configured to receive a communication from a widget, being software code residing in a website in response to a dealership website being loaded by a network device of a consumer, to obtain information pertaining to the vehicle including the modified image of the vehicle.
13. An image readjustment method, comprising:
receiving and analyzing an incoming image, using a machine learning analysis, to determine a location of a particular component within the incoming image and to calculate coordinates of a boundary area surrounding the particular component;
determining an aspect ratio of the incoming image;
generating a secondary boundary area based on both the aspect ratio determined for the incoming image and the coordinates of the boundary area, the secondary boundary area including the particular component substantially centered within the secondary boundary area and an aspect ratio that is (i) equal to the aspect ratio of the incoming image or (ii) substantially equivalent to the aspect ratio of the incoming image as being within three-percent of the aspect ratio of the incoming image; and
removing portions of the incoming image outside of the secondary boundary area to generate a modified image to be rendered for display.
14. The image readjustment method of claim 13 , wherein analyzing of the incoming image comprises calculating the coordinates of the boundary area and at least two coordinates of the boundary area for use in removing of the portions of the incoming image outside of the secondary boundary area.
15. The image readjustment method of claim 13 , wherein the modified image and coordinates of the particular component are stored within a persistent storage and accessed to render the modified image.
16. The image readjustment method of claim 13 , wherein the particular component is an image of a vehicle.
17. The image readjustment method of claim 16 further comprising:
receiving a communication from a widget, being software code residing in a website, to obtain information pertaining to the vehicle including the modified image of the vehicle.
18. The image readjustment method of claim 13 further comprising:
removing portions of the incoming image outside of the secondary boundary area including at least a portion of the boundary area.