API Documentation

Authentication

API Key

ps-api-k8f2m9x4n7j1w5...

Base URL

https://api.pixelsense.ai/v1

Include your API key in the Authorization header as Bearer <your-api-key>

Endpoints

POST/api/v1/analyze

Analyze a single image with specified pipeline and models.

Request Body

{ "image_url": "string", "pipeline": ["object_detection", "ocr"], "models": { "object_detection": "yolov8" }, "min_confidence": 0.7 }

Response

{ "id": "string", "detections": [...], "ocr": {...}, "models": [...], "totalTime": 1.2 }

Python Example

response = requests.post(
    "https://api.pixelsense.ai/v1/analyze",
    headers={"Authorization": "Bearer ps-api-k8f2m9..."},
    json={
        "image_url": "https://example.com/warehouse-photo.jpg",
        "pipeline": ["object_detection", "ocr"],
        "models": {
            "object_detection": "yolov8-custom-warehouse",
            "ocr": "gpt-4-vision"
        },
        "min_confidence": 0.7,
        "output": "json"
    }
)
result = response.json()
print(f"Objects detected: {len(result['detections'])}")

POST/api/v1/analyze/batch

Submit a batch of images for parallel analysis.

Request Body

{ "images": ["url1", "url2", ...], "pipeline": [...], "callback_url": "string" }

Response

{ "batch_id": "string", "status": "queued", "image_count": 156 }

POST/api/v1/analyze/url

Analyze an image directly from a public URL.

Request Body

{ "url": "https://example.com/image.jpg", "pipeline": ["object_detection"] }

Response

{ "id": "string", "detections": [...], "totalTime": 1.1 }

GET/api/v1/results

List all analysis results with pagination and filtering.

Response

{ "results": [...], "total": 48247, "page": 1, "limit": 20 }

GET/api/v1/results/:id

Get a specific analysis result with full detection data.

Response

{ "id": "string", "detections": [...], "ocr": {...}, "classification": [...] }

GET/api/v1/models

List all available models including pre-built and custom.

Response

{ "models": [...], "total": 8 }

POST/api/v1/models/train

Start training a custom model on a specified dataset.

Request Body

{ "dataset_id": "string", "base_model": "yolov8", "epochs": 100, "name": "my-custom-model" }

Response

{ "training_id": "string", "status": "started", "estimated_time": "2 hours" }

GET/api/v1/datasets

List all datasets with image counts and class distributions.

Response

{ "datasets": [...], "total": 3 }

POST/api/v1/datasets

Create a new dataset for custom model training.

Request Body

{ "name": "string", "description": "string", "split": { "train": 80, "validation": 10, "test": 10 } }

Response

{ "id": "string", "name": "string", "status": "created" }

GET/api/v1/analytics

Get usage analytics including volume, accuracy trends, and model performance.

Response

{ "processingVolume": [...], "analysesByType": [...], "modelAccuracy": [...] }

Loading analysis...

Endpoints

POST/api/v1/analyze

Analyze a single image with specified pipeline and models.

Request Body

{ "image_url": "string", "pipeline": ["object_detection", "ocr"], "models": { "object_detection": "yolov8" }, "min_confidence": 0.7 }

Response

{ "id": "string", "detections": [...], "ocr": {...}, "models": [...], "totalTime": 1.2 }

Python Example

response = requests.post(
    "https://api.pixelsense.ai/v1/analyze",
    headers={"Authorization": "Bearer ps-api-k8f2m9..."},
    json={
        "image_url": "https://example.com/warehouse-photo.jpg",
        "pipeline": ["object_detection", "ocr"],
        "models": {
            "object_detection": "yolov8-custom-warehouse",
            "ocr": "gpt-4-vision"
        },
        "min_confidence": 0.7,
        "output": "json"
    }
)
result = response.json()
print(f"Objects detected: {len(result['detections'])}")

POST/api/v1/analyze/batch

Submit a batch of images for parallel analysis.

Request Body

{ "images": ["url1", "url2", ...], "pipeline": [...], "callback_url": "string" }

Response

{ "batch_id": "string", "status": "queued", "image_count": 156 }

POST/api/v1/analyze/url

Analyze an image directly from a public URL.

Request Body

{ "url": "https://example.com/image.jpg", "pipeline": ["object_detection"] }

Response

{ "id": "string", "detections": [...], "totalTime": 1.1 }

GET/api/v1/results

List all analysis results with pagination and filtering.

Response

{ "results": [...], "total": 48247, "page": 1, "limit": 20 }

GET/api/v1/results/:id

Get a specific analysis result with full detection data.

Response

{ "id": "string", "detections": [...], "ocr": {...}, "classification": [...] }

GET/api/v1/models

List all available models including pre-built and custom.

Response

{ "models": [...], "total": 8 }

POST/api/v1/models/train

Start training a custom model on a specified dataset.

Request Body

{ "dataset_id": "string", "base_model": "yolov8", "epochs": 100, "name": "my-custom-model" }

Response

{ "training_id": "string", "status": "started", "estimated_time": "2 hours" }

GET/api/v1/datasets

List all datasets with image counts and class distributions.

Response

{ "datasets": [...], "total": 3 }

POST/api/v1/datasets

Create a new dataset for custom model training.

Request Body

{ "name": "string", "description": "string", "split": { "train": 80, "validation": 10, "test": 10 } }

Response

{ "id": "string", "name": "string", "status": "created" }

GET/api/v1/analytics

Get usage analytics including volume, accuracy trends, and model performance.

Response

{ "processingVolume": [...], "analysesByType": [...], "modelAccuracy": [...] }