Object Detection
- 0 Collaborators
By using Intel openVino and Intel Pyipp library doing a project on object detection and the DNN datasets are used for detecting over 80 objects in a single video or image. ...learn more
Project status: Under Development
Overview / Usage
Detecting an is classified by the of the pixels in the image. In the object detection the object is to be detected by its shape , color and the identification of the character . By using the object detection the machine will identify an object or something by given inputs in the datasets of YOLO weights and YOLO config or SSD_mobilenet_v3. Object detection models are usually trained to detect the presence of specific objects. The constructed models can be used in images, videos, or real-time operations. Even before the deep learning methodologies and modern-day image processing technologies, object detection had a high scope of interest.
Methodology / Approach
**Data Collection and Preparation : **Collect a dataset containing images or video frames with objects of interest. Annotate the dataset to label the objects and their locations within each image or frame.
Preprocessing :
Resize images to a consistent size. Normalize pixel values to ensure uniformity. Data augmentation techniques can be applied to increase the diversity of the dataset.
Selecting a Model Architecture:
Choose a deep learning architecture suitable for object detection.
- Faster R-CNN (Region-based Convolutional Neural Network)
- YOLO (You Only Look Once)
- SSD (Single Shot MultiBox Detector)
- Mask R-CNN (an extension of Faster R-CNN for instance segmentation)
Inference:
SSD (Single Shot MultiBox Detector) detection on new, unseen data. Pass the image or video frame through the model to obtain predictions. Decode the model's output to obtain bounding boxes and class labels for detected objects.
Visualization and Interpretation:
Overlay bounding boxes and class labels on the original images or frames to visualize detections . Interpret the results for downstream applications, such as tracking or decision-making.
Technologies Used
1.By using Intel openvino library to identify the object by using optimized computer vision.
2.Using Intel IPP library the resolution and some additional features are available in setting up the image or video to best format.
3.Using DNN trained like **YOLO weights_v7 and YOLO config_v7 or SSD_mobilenet_v3 **(Single Shot MultiBox Detector (SSD) etc.
4.For plotting process in the image detection matplotlib is plays a vital role in the detection it is used to mark and capture the pixels in the image.
5.Devcloud is used to run the program.