Java Image Similarity Search App Development
Budget: €8 – €30 EUR
I'm looking for a skilled Java developer to construct an Image Similarity Search application that operates on a Client-Server architecture and relies on a database for image indexing. This application will mainly target the Windows platform.
Requirement 1: Initialize a set of images present on the server
Description: When the server is launched, it must charge all the images present at the initialization of the software as well as their index in a base of data (see Requirement 3). The folder where images must exist at the server launch. Only the corresponding image sub-folders need to be processed. The sub-folders of the folder images that are not processed.
Error Handling: If the image does not exist where the server was launched, an error must be thrown.
Tests:
Launching the executable from an empty environment, an error should trigger indicating the image folder is not present.
Place one or several image files with recognized formats as well as documents with unrecognized extensions being image files (e.g., .txt).
Requirement 2: Generate images present on the server
Description: The server can generate thumbnails, i.e., produce the raw data of each image as well as the necessary metadata for responses to requests (identifier, name of the file, image type, format...). The server can:
access an image via its identifier,
upload an image via its identifier,
delete an image,
construct the list of available images (composed uniquely of metadata).
The server guarantees the images folder: the images content folder at any moment contains the list of available images that are also indexed in the database.
Requirement 3: Index an image
Description: The server allows indexing an image that is to say to add into the database of corresponding data the identifier (identifier and descriptor). The supported descriptors are the 2D histogram Texture/Saturation and the 3D RGB histogram of the image.
Requirement 4: Search for images similar to a given image
Description: The server can construct the list of N images most similar to a given image based on a given descriptor.
Ideal candidates should have:
- Strong Java programming skills
- Extensive experience with PostgreSQL
- Experience with Image Similarity Search algorithms
This project is a fantastic opportunity for any developer looking to showcase their skills in a value-driven environment. Ideal candidates should be results-driven and ready to hit the ground running. Let's make something great together.
Requirement 1: Initialize a set of images present on the server
Description: When the server is launched, it must charge all the images present at the initialization of the software as well as their index in a base of data (see Requirement 3). The folder where images must exist at the server launch. Only the corresponding image sub-folders need to be processed. The sub-folders of the folder images that are not processed.
Error Handling: If the image does not exist where the server was launched, an error must be thrown.
Tests:
Launching the executable from an empty environment, an error should trigger indicating the image folder is not present.
Place one or several image files with recognized formats as well as documents with unrecognized extensions being image files (e.g., .txt).
Requirement 2: Generate images present on the server
Description: The server can generate thumbnails, i.e., produce the raw data of each image as well as the necessary metadata for responses to requests (identifier, name of the file, image type, format...). The server can:
access an image via its identifier,
upload an image via its identifier,
delete an image,
construct the list of available images (composed uniquely of metadata).
The server guarantees the images folder: the images content folder at any moment contains the list of available images that are also indexed in the database.
Requirement 3: Index an image
Description: The server allows indexing an image that is to say to add into the database of corresponding data the identifier (identifier and descriptor). The supported descriptors are the 2D histogram Texture/Saturation and the 3D RGB histogram of the image.
Requirement 4: Search for images similar to a given image
Description: The server can construct the list of N images most similar to a given image based on a given descriptor.
Ideal candidates should have:
- Strong Java programming skills
- Extensive experience with PostgreSQL
- Experience with Image Similarity Search algorithms
This project is a fantastic opportunity for any developer looking to showcase their skills in a value-driven environment. Ideal candidates should be results-driven and ready to hit the ground running. Let's make something great together.