40 applications of Halcon software in image processing and machine vision

Identifying Different Models

We can identify different models from each other using matching algorithms (object recognition). As shown in the image, three different pieces are distinguished from each other, each displayed with a different color.

Traffic Monitoring with Image Processing

Today, in advanced vehicle surveillance systems, we use image processing to detect moving vehicles and individuals. This way, images captured from the video are analyzed simultaneously with video recording by cameras, and vehicles and individuals are accurately identified.

Product Packaging Inspection

Using image processing, we can ensure the completeness of the desired object. To illustrate this, pay attention to the highlighted areas in the upper figure, where each package includes various flavors of cheese, and with image processing, we can ensure that all these flavors are included in the package, or not?

Quality Control of Tablets

Blister means a bubble and is one of the best packaging methods for pharmaceuticals and food items. With image processing, we can identify fractures, cracks, and color differences in various types of tablets and capsules.

Plant Identification with Image Processing

With image processing, we can identify plants in different backgrounds and recognize their type. As an example in the image above, a plant has been identified and highlighted with green color through image processing.

Object Motion Tracking

With image processing, we can identify objects in motion. The image taken in motion shows the motion tracking of a motor, performed by image processing, and its position in the image is determined.

2D Measurement with Image Processing

In 2D measurement, geometric shapes such as circles, ellipses, rectangles, or part of an arc can be identified using edge detection in the image, and parameters such as the radius can be extracted.

Road Signs Identification

With image processing, we can detect and recognize road signs and guidance during vehicle movement. In this case, if the driver fails to notice the signs due to lack of sufficient concentration or inattention, the vehicle monitoring system equipped with image processing can alert him and prevent potential hazards. The image above is taken by the monitoring system of a moving vehicle, where a warning sign has been detected.

Label Position Control

With image processing, we can check the accuracy of label position. As shown in the images, the label on the shampoo must be placed in a specific position on the shampoo with a slight tolerance. Using image processing, we detect this tolerance to take corrective actions in case of deviation from the allowable amount.

One-Dimensional Measurement

Using image processing, the distance between two points can be determined within seconds using filter-based detection. In the figure, the distance between the blades of a fan is obtained using image processing. The red color indicates that this distance has exceeded the permissible tolerance, meaning it is either greater or less than the allowable limit.

Road Identification

Road identification from aerial photos is achievable through image processing. An enlarged aerial photo in the figure shows a road identified using image processing.

Surface Defect Detection with Image Processing

As shown in the images, surface defects on plastic surfaces are identified and the defective area is marked with a red circle using image processing.

Motion Detection with Optical Flow

Optical flow is a method for motion detection in images, where two consecutive images from a camera are compared to identify the position, speed, and direction of motion. In the example, bicycle movement is detected using image processing with Halcon software, triggering the opening command of the jack.

Product Fault Detection with Image Processing

Using image processing algorithms, cracks and bumps on the surface of battery lids can be identified. The image shows examples of detected faults on the surface of a battery lid.

Calculating the Lifespan of Trees

The simplest way to calculate the lifespan of trees is by counting their layers, with each less colorful and colorful layer representing one year of the tree’s life. Using image processing, we can count these layers and calculate the lifespan of the tree.

3D Object Detection

With image processing, we can find the coordinates of three-dimensional parts. These coordinates can be used for robot gripping, counting, part identification, etc. The figure shows the identified position of a three-dimensional part using image processing.

Texture Identification with Image Processing

Using image processing, we can identify textures, a process known as Texture Inspection. Texture identification allows us to detect defects and irregularities in the texture. As seen in the above images, using texture identification in image processing, the area of surface irregularity is clearly identified.

Color Identification with Image Processing

With image processing, we can identify colors. Color identification in image processing greatly aids in various algorithms. In the example, fuse colors are identified using image processing, and their color determines the allowable fuse current (e.g., a fuse with blue color can tolerate a current of 15 amps).

Creating Mosaic Images with Image Processing

Creating mosaic images involves connecting images from different parts of a scene to obtain a larger seamless image. Mosaic methods can be used to combine satellite images to achieve a larger area. The above figure shows six separate images merged together using mosaic techniques.