Monitoring the harvesting and processing of fruit, vegetables and crops is essential to obtain qualitive and safe foods. Computer vision applications during harvesting, cleaning and sorting processes ensure pure, ripe and healthy products in the proper shapes and desired sizes.
APIXA applies next-generation cameras and hardware combined with AI, and possibly deep learning, to judge on crystal clear images and take the right decisions. Fruit, vegetables and crops are inspected on conveyor belts or when falling using 2D monochrome or RGB, 3D, multispectral or hyperspectral cameras. Internal defects can be traced with an x-ray source. In certain applications, contaminations in the form of mildew or bird droppings are detected using drones to avoid disease outbreaks and ensure pure and clean products. Also sorting applications concerning fruits, such as cacao beans and almond nuts, and processed foods, such as fries and bakery products, benefit greatly from smart computer vision applications.
The APIXA hardware and software solutions enable producers and processors to detect anomalies and inspect product composition, sort and grade products towards size and shape, and avoid contaminations and disease outbreaks. In the process, the trainable solutions are designed to continuously improve performance robustness and exclude false positives. APIXA is a pioneer in the use of hyperspectral cameras, deep learning algorithms and vision-guided robotics.
Crop mildew detection
Detecting mildew on a variety of fruits and vegetables is critical to prevent the spread of this crop disease. Drones equipped with hyperspectral cameras and AI driven data processing identify infected crops in an early stage so that prompt corrective action can be taken. This avoids large crop losses on plantations with lettuce, beets, spinach, grapevines, etc.
Premium nut quality
More expensive nuts, such as almonds, pistachios and walnuts, are inspected for grading purposes. APIXA applies multiple 3D RGB cameras and AI data processing to identify defect levels and evaluate quality related aspects of falling nuts. The nut bulk price is determined by distinguishing the shape and size characteristics, insects inflicted damage, structural uniformity, product integrity, and other aspects. Computer vision outperforms manual grading in terms of productivity, objectivity and level of detail.
Quality inspection of fries
The French fries industry is big business and premium brands aim at offering top quality fries. Based on single sensor (RGB or hyperspectral imaging) or sensor fusion imaging, APIXA applies AI data processing algorithms to segment individual fries in images taken on the conveyor belt of sorting infrastructure. This segmenting allows for rejecting fries that are out of specification in terms of size, thickness, color or uniformity.
Cacao beans composition
The quality of cacao beans, with pods removed, can be determined through hyperspectral imaging. This type of chemical imaging provides graphical insights into the chemical composition of the beans with accuracy far beyond human eyesight.
X-ray density imaging
Hard objects in food are undesirable. X-rays visually distinguish pits from cherries, shells from almond nuts, or even loose shell flakes from scallops. Also, internal structural damage can be detected in this way. APIXA develops software for item segmentation and category classification using deep learning. Compared to classical vision methods, APIXA's deep learning algorithms perform significantly better and faster. Furthermore, the algorithms are generic in nature and therefore eliminate the lengthy and costly process of developing a new specific algorithm for each new food type.
Food packing verification
Detecting defective food packaging is important for manufacturers to maintain premium quality and a solid reputation. APIXA computer vision and AI data processing detects out-of-sync cutting of rice cracker packaging. As a result, the robust cutting of the plastic flowpack wrappings avoids shelf-life problems and product complaints.
Potato chips inspection
Manufacturers prefer to have potato chips with holes and black spots removed before entering the chips bag. Superfast 2D images of structured chips flakes taken from above the conveyor belts are analyzed centrally using performant AI algorithms. These execute the segmentation, positioning, orientation and defect recognition of individual potato chips.
Vision-guided robotics are increasingly used to optimize the process of planting seedlings. Computer vision and AI data processing can be used to increase the planting precision and success rate for planting lettuce seedlings, for example. APIXA specialists determine what type of imaging and AI algorithms suit best your food related application. Typically, fast imaging and reliable robot feedback are essential in vision-guided robotics.