SICK Application | SICK Inspector8512 Smart Camera + AI Powers Candy Residue Inspection on Molds

Inspection of candy residue on molds is a crucial quality control step after candy production. Traditional vision algorithms feature cumbersome debugging and high deployment costs. Equipped with high-resolution imaging, flexible integration capability and powerful built-in AI inspection tools, the SICK Inspector8512 smart camera delivers a stable, accurate and professional solution for candy residue detection on molds.

Candy residue inspection on molds is vital to guarantee product quality, demolding effect and food safety during candy manufacturing. Nonetheless, special mold structures and harsh production line conditions bring multiple challenges to conventional vision algorithms:

1. Diverse shapes of products and molds

   Complex mold structures and various candy types result in poor versatility of traditional algorithms. They fail to adapt to diverse scenarios, accompanied by tedious debugging and high adaptation costs.

2. Low distinguishability due to similar color and texture

   Candies share similar color with mold materials with low contrast. Conventional vision algorithms are prone to misjudgment and missing detection, making stable identification of residual candies impossible.

3. Insufficient equipment integration and compatibility

   Multiple industrial communication protocols need to be supported, while traditional devices have limited communication performance. Complicated system configuration leads to difficult and costly deployment and maintenance.

4. Strict on-site reliability requirements

   Inspection results directly affect quality management and production traceability. Changes of workshop temperature, humidity, light and other environmental factors easily undermine the stability of traditional solutions.

1. Detect randomly distributed residual candies on molds.
2. Transmit detection results to PLC via PROFINET communication.

1. The field of view is wide, while residual candies to be detected are tiny, forming small target detection within a large viewing range.
2. Residual candies vary in color and size.

1. After positioning the camera and light source, access the camera software interface via browser. Set proper parameters including exposure time, focal length, contrast and brightness to capture clear images.

Clarity image

Add the AI object detection tool and pre-train the model with samples of residual candies at different positions on molds to identify candy residues on mold surfaces. The Inspector8512 smart camera is equipped with Nova image processing software, which integrates abundant image processing algorithms for users to quickly build customized processing workflows.

Comprehensive image processing algorithms

No OK image of the candy was detected.

Detected the NG image of the candy

Transmit detection results to PLC or host computer via applicable communication modes, including digital IO output signals, TCP/IP, PROFINET and EtherNet/IP. Captured images can also be saved to the host computer through FTP for subsequent traceability and inquiry.

1. 12-megapixel high-resolution imaging: Equipped with a 12MP high-definition CMOS sensor and external high-performance lighting, it clearly captures residual candies in mold gaps and edges.

2. Minimalist web visual configuration: Adopting web interface based on SICK Nova platform. No professional programming required, enabling both technical and general staff to quickly set parameters and build inspection solutions.

3. AI intelligent detection: Adopt AI object detection tool. Accurately identify mold candy residues via sample training without complex rule programming, delivering stable and reliable inspection performance.

4. Flexible optical adaptation: Standard C-Mount lens interface with manually adjustable focal length. Compatible with various external lenses and lighting modules to meet diverse mold installation and inspection requirements.

5. Easy industrial integration: Supports mainstream industrial buses including dual-port EtherNet/IP and PROFINET. Combined with high-speed I/O, it can be rapidly connected to production line PLC and control systems.